Abstract Aims Vulnerable coronary plaque is characterized by a large lipid core. Although commercially-available optical coherence tomography (OCT) systems use near-infrared light at 1300 nm wavelength, lipid shows characteristic absorption at 1700 nm. Therefore, we developed a novel, short wavelength infra-red, spectroscopic, spectral-domain OCT. The aim of the present study is to evaluate the accuracy of short wavelength (1700 nm) infra-red optical coherence tomography (SWIR-OCT) for identification of lipid tissue within coronary plaques. Methods and results Twenty-three coronary arteries from 10 cadavers were imaged at physiological pressure with 2.7 Fr SWIR-OCT catheter. When a blood-free image was observed, the SWIR-OCT imaging core was withdrawn at a rate of 20 mm/s using an automatic pullback device. SWIR-OCT images were acquired at 94 frames/s and digitally archived. SWIR-OCT generated grey-scale cross sectional images and colour tissue maps of all of the plaque by using a lipid analysis algorithm. After SWIR-OCT imaging, the arteries were pressure-fixed, sliced by cryostat and stained with Oil Red O, and then corresponding histology was collected in matched images. Regions of interest, selected from histology, were 117 lipidic and 34 fibrotic/calcified regions. SWIR-OCT showed high sensitivity (89%) and specificity (92%) for identifying lipid tissue within coronary plaques. The positive predictive value and negative predictive value were 97% and 74%, respectively. Conclusion SWIR-OCT accurately identified lipid tissue in coronary autopsy specimens. This new technique may hold promise for identifying histopathological features of coronary plaque at risk for rupture. optical coherence tomography , lipid-rich plaque , short wavelength infra-red , spectrum Introduction A large lipid core can cause acute coronary syndromes and cannot be accurately detected by conventional diagnostic methods. An intracoronary imaging device that could detect lipid tissue could be used for clinical risk stratification and as a guide to therapeutic decision during percutaneous coronary intervention (PCI). Although conventional imaging modalities, such as intravascular ultrasound (IVUS) and optical coherence tomography (OCT), are well correlated with the histological features, diagnostic accuracy may depend on the proficiency of the physician.1,2 Advanced intracoronary imaging devices, such as colour-coded IVUS and near-infrared spectroscopy intravascular ultrasound (NIRS-IVUS), were developed to obtain plaque composition information.3,4 Although these devices are useful for estimating plaque composition, the cross-sectional images are low resolution compared to OCT. Although colour-coded OCT such as polarization sensitive-optical coherence tomography has been in development for years, it is not yet commercially available. We developed a new colour-coded OCT system for lipid detection. In the OCT system, as light passes through the tissue, it is attenuated by scattering and absorption. Scattering is the predominant form of attenuation encountered by conventional OCT systems that have wavelengths near 1300 nm. As highly attenuating tissue, such as lipid, has low penetration depth, OCT does not allow a deep view of some lipid-containing plaques, despite its high resolution.5,6 Tangential drop out signs, an example of OCT artefacts, also appear, although the signal is poor.7 It is difficult to discriminate between a lipid plaque and an artefact by OCT. Short wavelength infra-red (SWIR) is a part of infrared wavelength, ranging between 1400 and 3000 nm, where molecular bonds, such as carbon–hydrogen bond, have characteristic attenuation spectrum. It is appropriate to use 1700 nm wavelength for lipid detection, because lipid shows characteristic absorption at 1700 nm in the SWIR range (Figure 1).8 We therefore developed a novel SWIR-OCT that can automatically detect lipid tissue using 1700 nm wavelength in contrast to conventional OCT using 1300 nm. This study aims to validate the lipid tissue identification ability in human coronary autopsy specimens. Figure 1 View largeDownload slide Measurement of light attenuation spectra of normal and lipid region in 1300 and 1700 nm wavelength bands. Four lipid (red circles) and four normal points (blue) were chosen in an unstained coronary frozen section (A), referring to the Oil Red O stained section (B). Light attenuation spectra were measured for the chosen points in conventional 1300 nm near-infrared (C) and 1700 nm SWIR (D) wavelength bands using a commercial spectrometer. Lipid has characteristic attenuation peak in SWIR. Figure 1 View largeDownload slide Measurement of light attenuation spectra of normal and lipid region in 1300 and 1700 nm wavelength bands. Four lipid (red circles) and four normal points (blue) were chosen in an unstained coronary frozen section (A), referring to the Oil Red O stained section (B). Light attenuation spectra were measured for the chosen points in conventional 1300 nm near-infrared (C) and 1700 nm SWIR (D) wavelength bands using a commercial spectrometer. Lipid has characteristic attenuation peak in SWIR. Methods Specimens Between December 2013 and August 2014, we examined 25 coronary arteries from 10 cadavers (9 male and 1 female, mean age 76 ± 9). The cause of death was ischaemic heart disease in 2 of the 10. Three had a prior history of ischaemic heart disease. Three of the ten were type 2 diabetes mellitus. The study protocol was approved by the Wakayama Medical University Ethics Committee, written informed consent was obtained from each family. OCT images were recorded within 24 h of hearts being obtained from cadavers at autopsy. The time between death and OCT imaging did not exceed 72 h and hearts were kept at 4˚C before OCT imaging. SWIR-OCT SWIR-OCT images were obtained by a prototype SWIR-OCT system (Sumitomo Electric Industries, Japan) as previously reported, with minor modifications (Figure 2).9,10 Briefly, the SWIR-OCT device was made of 1700 nm superluminescent diode light source and SWIR spectroscopic cameras. Coronary arteries were cut at 4 cm or more from the ostium, and 2.7 Fr OCT imaging catheters were inserted from the cutting site to the coronary ostium with guide wires. The imaging lens in the OCT catheter was pulled-back at 20 mm/s by an automatic pull-back device during continuous infusion of phosphate-buffered saline perfused at 100 mmHg using a pressure infuser. For off analysis, the SWIR-OCT images were acquired at 94 frames/s and digitally archived in the OCT system console. OCT raw data was spectrally analysed by proprietary software to generate lipid-enhanced OCT images. Figure 2. View largeDownload slide SWIR-OCT system. Prototype SWIR-OCT system (Sumitomo Electric Industries, Japan) (A) and software flow chart (B). The system measures spectrum of the artery under test at 1700 nm wavelength band. Fourier analysis on the spectrum generates a standard OCT image and a spectral analysis generates a lipid distribution image, which is superimposed on the OCT image, resulting in a lipid-enhanced OCT image. Figure 2. View largeDownload slide SWIR-OCT system. Prototype SWIR-OCT system (Sumitomo Electric Industries, Japan) (A) and software flow chart (B). The system measures spectrum of the artery under test at 1700 nm wavelength band. Fourier analysis on the spectrum generates a standard OCT image and a spectral analysis generates a lipid distribution image, which is superimposed on the OCT image, resulting in a lipid-enhanced OCT image. Histological examination After OCT imaging, we applied pressure-fixation to maintain orientation and size for comparison with the OCT images as previously reported.11 Briefly, the ascending aorta and pulmonary artery were excised at about 3–5 cm downstream from the aortic and pulmonary valves, respectively and cannulated with a 10 cm long, rigid, plastic cylinder. The other end of the plastic cylinder was connected to a bag filled with formalin via flexible plastic tubing. The ends of cut coronary arteries were ligated with sutures. The aorta and pulmonary arteries were perfused with 10% neutral buffered formalin by establishing a closed system between the heart and the bag. Pressure inside the left and right heart system was maintained at a physiological level (100 and 30 mmHg, respectively) with a pressurizer. After fixation for a week, epicardial arteries were removed from the heart and decalcified with ethylenediaminetetraacetic acid. The proximal and middle segments of the three arteries were examined. The left main coronary artery, distal segments and all branches were excluded from the analysis. Dissected segments were serially sectioned to the longitudinal axis of the vessel at 3 mm intervals, and then transferred to 20% sucrose in phosphate-buffered saline overnight at 4 °C. After embedding in OCT compound (Sakura Finetek USA, Torrance, CA, USA), 6 µm cross-sections of the arteries were mounted on slides and stained with haematoxylin and eosin (HE) and Oil red O. Each histopathologic slide was digitized using a microscope at low magnification (×1.25). A total of 386 sections were obtained from the autopsy specimens. Of those, 235 sections were excluded because of visually normal coronary arteries (n = 220) or because of cutting artefacts (n = 15). Finally, a total of 10 cadavers with 151 sections were examined. Correlation with SWIR-OCT image and histology and plaque differentiation Every histopathologic slice and its corresponding OCT image was matched using vessel shape, side branches, perivascular structures, and distances from side branches or the cut distal end.12 All plaques were checked by a pathologist who was unaware of the OCT results and classified as one of three types according to AHA classification: calcification predominates (calcified plaque), confluent extracellular lipid core formed/fibromuscular tissue layers produced (lipid plaque), or fibrous tissue predominates (fibrous plaque).13 OCT images were also reviewed by experienced observers, unaware of the histologic examination data and using established criteria, and again classified as either calcified, lipid, or fibrous plaque.1,5,14 Lipid detection by SWIR-OCT and histology SWIR-OCT software analysed a light spectrum of each point within a cross sectional SWIR-OCT image. Any point characteristic of the lipid spectrum was classified as a lipid and marked in yellow. Non-lipid regions (normal tissue, calcification, etc.) were marked in green. When a region did not contain enough signals to be analysed, it was not classified and no colour was assigned. A lipid image of the entire cross section was generated by colouring all the points in the SWIR-OCT image. A lipid-enhanced OCT image was then reconstructed in SWIR-OCT software by superimposing the lipid image on the standard OCT image (Figure 2). Frozen sections were used for HE and Oil Red O stains, which stained lipid tissue red. Lipid tissue was defined when the tissue stained red by Oil Red O and was not apparently calcified, as judged by HE stain. A section stained by Oil red O was quantitatively compared with the corresponding lipid-enhanced OCT image as a lipid detection analysis. Statistical analysis Statistical analyses were performed by using the R statistical package, version 3.1.0 (The R Foundation for Statistical Computing). The degree of agreement between the histologic diagnosis and the results obtained by the SWIR-OCT readers was quantified by the Cohen’s kappa test of concordance. A value of 0.61–0.80 indicates good agreement, and 0.81–1.0 indicates excellent agreement. Results We examined 151 cross-sections in 14 coronary arteries from 10 cadavers (4 left anterior descending, 4 left circumflex, and 6 right coronary arteries). All sections were well matched between histology and SWIR-OCT images and analysed for the SWIR-OCT accuracy assessment. Overall agreement between the OCT and the histological diagnoses of plaque types was good (Cohen’s ĸ = 0.74, 95% CI: 0.64–0.83). On either a histological section, or the matched SWIR-OCT image, 117 sections were lipid composition. Representative OCT image and lipid-enhanced OCT images are shown in Figure 3A and B. The lipid-enhanced OCT image showed lipid by yellow colouring, which almost exactly coincided with the area stained by Oil Red O (Figure 3D). Test results of lipid detection analysis on 151 sections were 89% sensitivity, 92% specificity, 97% positive predictive value, and 74% negative predictive value (Tables 1 and 2). Furthermore, sensitivity, specificity, positive predictive value, and negative predictive value of plaque classification analysis with normal SWIR-OCT using histology as gold standards were 89%, 92%, 80%, and 96%, respectively. Table 1 Assessment of SWIR-OCT for lipid detection Histology Total Lipid (+) Lipid (−) SWIR-OCT Lipid (+) 101 3 104 Lipid (–) 12 35 47 Total 113 38 151 Histology Total Lipid (+) Lipid (−) SWIR-OCT Lipid (+) 101 3 104 Lipid (–) 12 35 47 Total 113 38 151 Table 1 Assessment of SWIR-OCT for lipid detection Histology Total Lipid (+) Lipid (−) SWIR-OCT Lipid (+) 101 3 104 Lipid (–) 12 35 47 Total 113 38 151 Histology Total Lipid (+) Lipid (−) SWIR-OCT Lipid (+) 101 3 104 Lipid (–) 12 35 47 Total 113 38 151 Table 2 Accuracy of SWIR-OCT for lipid detection compared with histology Sensitivity (%) 89 Specificity (%) 92 Positive predictive value (%) 97 Negative predictive value (%) 74 Sensitivity (%) 89 Specificity (%) 92 Positive predictive value (%) 97 Negative predictive value (%) 74 Table 2 Accuracy of SWIR-OCT for lipid detection compared with histology Sensitivity (%) 89 Specificity (%) 92 Positive predictive value (%) 97 Negative predictive value (%) 74 Sensitivity (%) 89 Specificity (%) 92 Positive predictive value (%) 97 Negative predictive value (%) 74 Figure 3 View largeDownload slide Representative SWIR-OCT image and corresponding histology. Representative SWIR-OCT image (A), lipid-enhanced OCT image (B), the corresponding HE stain (C), and Oil Red O stain (D). Lipid-enhanced image showed lipid with yellow colour, which almost coincided with Oil Red O. All scale bars represent 1 mm. Figure 3 View largeDownload slide Representative SWIR-OCT image and corresponding histology. Representative SWIR-OCT image (A), lipid-enhanced OCT image (B), the corresponding HE stain (C), and Oil Red O stain (D). Lipid-enhanced image showed lipid with yellow colour, which almost coincided with Oil Red O. All scale bars represent 1 mm. Discussion We developed a SWIR-OCT system that can automatically detect lipid tissue. The present study shows that this SWIR-OCT system can identify lipid tissue in human coronary artery autopsy specimens as validated by histology with high accuracy. Detection of lipid tissue of SWIR-OCT using histology as gold standard was 90% accurate. This value is representative of the performance of commonly used intracoronary modalities.4,5,15 In detecting lipid tissue, the sensitivity, specificity, positive predictive value and negative predictive value were 89%, 92%, 97%, and 74%, respectively. The low negative predictive value was expected due to small number and the attenuation or loss of penetration depth of OCT signal, which was excluded from lipid detection analysis by SWIR-OCT. The present study is similar to ex vivo studies of other intravascular imaging devices, such as conventional OCT, IVUS, and near infra-red spectroscopy (NIRS), in that we validated atherosclerotic plaque compared with the histology.1,4,5,16,17 Our study differs in that the frozen sections were used for the histological analysis. Lipid composition is eluted during the preparation of paraffin blocks, whereas frozen sections can maintain lipid composition. We could thereby detect lipid region more accurately in frozen than in paraffin sections. Conventional OCT light is in the near infra-red range, typically with wavelengths around 1300 nm for two reasons: to obtain clear images because blood has very low absorption at this wavelength and to obtain penetration depth (shorter being better).18 Conversely, SWIR-OCT with wavelength around 1700 nm is potentially inferior to conventional OCT for clarity of images and penetration depth. In order to solve the image clarity problem, the width of frequency range of SWIR-OCT was designed to have spatial resolution as fine as conventional OCT with regard to the penetration depth, determined by optical absorption and scattering. 1700 nm is inferior for absorption, but better for scattering when compared to 1300 nm. It was reported that the main limitation of penetration depth of OCT was optical scattering rather than absorption, however.19 As the present OCT system used a camera for acquiring spectrum, the spatial resolution was unfortunately not good. We plan to measure the spectrum using a laser to demonstrate that SWIR-OCT is not inferior to conventional OCT in either spatial resolution or penetration. The SWIR-OCT system has some important clinical utilities. Whereas the interpretation of OCT images requires a lot of skill and training, SWIR-OCT overcomes this problem by detecting lipid plague without particular necessity for a highly skilled OCT reader. Moreover, this lipid detecting function of SWIR-OCT is useful for distinguishing lipid plaque and tangential drop sign from signal poor regions because its lipid detecting system analyses the characteristic spectrum from lipid compositions. Lipid identification might help in making a decision to stent during coronary intervention, as both plaque structure and morphology are highly informative with regard to vulnerability. SWIR-OCT also has the potential to identify calcified and fibrous tissue in colour mapping. Although overlap between lipid and calcified regions in frozen sections were observed, this present study focuses on only lipid detection. We recognize the need for improvement in identifying both lipid and calcified composition on coloured SWIR-OCT images. Additional studies are also planned to validate the safety of using imaging catheters and the ability of the technique to identify tissue type in the coronary artery in vivo. Study limitations We assessed coronary arteries from cadavers. However, because the specimens were maintained at 4˚C until studied, the OCT images obtained here were unlikely to differ significantly from in vivo intracoronary OCT images.20,21 Due to the difficulty in obtaining human autopsy samples, the present study utilized a limited number of cases. Further studies are required to validate the relevance and application of our findings in a larger cohort. Another potential limitation is that, although our interpretation of lipid tissue was analysed according to unstained frozen sections, the effect of lipid degradation while embedding is unknown. Finally, the penetration depth of the optical coherence tomographic signal is limited to 2 mm, this limited penetration depth presents difficulties in spectrum analysis by SWIR-OCT at deep tissue sites. Conclusions We developed a novel SWIR-OCT system to automatically detect lipid tissue. SWIR-OCT accurately identified lipid tissue in human coronary autopsy specimens. This new technique holds promise for identifying histopathological features of coronary plaque at risk for rupture. Acknowledgements The authors would like to thank Mayumi Akira and Masako Nakanishi for technical assistance, and Yasuteru Muragaki, Department of Pathology, Wakayama Medical University, Japan, for his valuable advice on the pathological aspects of the work. Funding This study was funded by Sumitomo Electric Industries, Ltd. and Wakayama Medical University. Conflict of interest: T.A. has received a research fund from Sumitomo Electric Industries, Ltd. Yokohama, Japan (grant no. G26008). A.S. has received a research fund from Wakayama Medical University (2013 Wakayama Medical Award for Young Researchers). All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. References 1 Yabushita H , Bouma BE , Houser SL , Aretz HT , Jang IK , Schlendorf KH et al. Characterization of human atherosclerosis by optical coherence tomography . Circulation 2002 ; 106 : 1640 – 5 . Google Scholar CrossRef Search ADS PubMed 2 Roelandt JR , Serruys PW , Bom N , Gussenhoven WG , Lancee CT , ten Hoff H. Intravascular real-time, two-dimensional echocardiography . Int J Cardiac Imag 1989 ; 4 : 63 – 7 . http://dx.doi.org/10.1007/BF01795127 Google Scholar CrossRef Search ADS 3 Garcia-Garcia HM , Gogas BD , Serruys PW , Bruining N. IVUS-based imaging modalities for tissue characterization: similarities and differences . Int J Cardiovasc Imaging 2011 ; 27 : 215 – 24 . Google Scholar CrossRef Search ADS PubMed 4 Gardner CM , Tan H , Hull EL , Lisauskas JB , Sum ST , Meese TM et al. Detection of lipid core coronary plaques in autopsy specimens with a novel catheter-based near-infrared spectroscopy system . JACC Cardiovasc Imaging 2008 ; 1 : 638 – 48 . Google Scholar CrossRef Search ADS PubMed 5 Kume T , Akasaka T , Kawamoto T , Watanabe N , Toyota E , Neishi Y et al. Assessment of coronary arterial plaque by optical coherence tomography . Am J Cardiol 2006 ; 97 : 1172 – 5 . Google Scholar CrossRef Search ADS PubMed 6 Tearney GJ , Regar E , Akasaka T , Adriaenssens T , Barlis P , Bezerra HG et al. Consensus standards for acquisition, measurement, and reporting of intravascular optical coherence tomography studies: a report from the International Working Group for Intravascular Optical Coherence Tomography Standardization and Validation . J Am Coll Cardiol 2012 ; 59 : 1058 – 72 . Google Scholar CrossRef Search ADS PubMed 7 van Soest G , Regar E , Goderie TP , Gonzalo N , Koljenovic S , van Leenders GJ et al. Pitfalls in plaque characterization by OCT: image artifacts in native coronary arteries . JACC Cardiovasc Imaging 2011 ; 4 : 810 – 3 . Google Scholar CrossRef Search ADS PubMed 8 Tanaka M , Hirano M , Hasegawa T , Sogawa I. Lipid distribution imaging in in-vitro artery model by 1.7-μm spectroscopic spectral-domain optical coherence tomography . Proc SPIE 2014 ; 8565 : 85654F . 9 Tanaka M , Okuno T , Obi H , Hattori I , Hirano M , Ueno T. Performance improvement by a broadband super-luminescent diode light source in 1.7-μm spectroscopic spectral-domain optical coherence tomography for lipid distribution imaging in a coronary artery . Proc SPIE 2014 ; 8926 : 89262T . 10 Hirano M , Tonosaki S , Ueno T , Tanaka M , Hasegawa T. Improved method to visualize lipid distribution within arterial vessel walls by 1.7 μm spectroscopic spectral-domain optical coherence tomography . Proc SPIE 2014 ; 8935 : 893517 . 11 Asakura T , Karino T. Flow patterns and spatial distribution of atherosclerotic lesions in human coronary arteries . Circ Res 1990 ; 66 : 1045 – 66 . http://dx.doi.org/10.1161/01.RES.66.4.1045 Google Scholar CrossRef Search ADS PubMed 12 Lee JB , Mintz GS , Lisauskas JB , Biro SG , Pu J , Sum ST et al. Histopathologic validation of the intravascular ultrasound diagnosis of calcified coronary artery nodules . Am J Cardiol 2011 ; 108 : 1547 – 51 . Google Scholar CrossRef Search ADS PubMed 13 Stary HC. Natural history and histological classification of atherosclerotic lesions: an update . Arterioscler Thromb Vasc Biol 2000 ; 20 : 1177 – 8 . http://dx.doi.org/10.1161/01.ATV.20.5.1177 Google Scholar CrossRef Search ADS PubMed 14 Xu C , Schmitt JM , Carlier SG , Virmani R. Characterization of atherosclerosis plaques by measuring both backscattering and attenuation coefficients in optical coherence tomography . J Biomed Opt 2008 ; 13 : 034003. http://dx.doi.org/10.1117/1.2927464 Google Scholar CrossRef Search ADS PubMed 15 Kawasaki M , Bouma BE , Bressner J , Houser SL , Nadkarni SK , MacNeill BD et al. Diagnostic accuracy of optical coherence tomography and integrated backscatter intravascular ultrasound images for tissue characterization of human coronary plaques . J Am Coll Cardiol 2006 ; 48 : 81 – 8 . Google Scholar CrossRef Search ADS PubMed 16 Moreno PR , Lodder RA , Purushothaman KR , Charash WE , O’Connor WN , Muller JE. Detection of lipid pool, thin fibrous cap, and inflammatory cells in human aortic atherosclerotic plaques by near-infrared spectroscopy . Circulation 2002 ; 105 : 923 – 7 . Google Scholar CrossRef Search ADS PubMed 17 Nair A , Kuban BD , Tuzcu EM , Schoenhagen P , Nissen SE , Vince DG. Coronary plaque classification with intravascular ultrasound radiofrequency data analysis . Circulation 2002 ; 106 : 2200 – 6 . Google Scholar CrossRef Search ADS PubMed 18 Brezinski ME , Tearney GJ , Bouma BE , Izatt JA , Hee MR , Swanson EA et al. Optical coherence tomography for optical biopsy. Properties and demonstration of vascular pathology . Circulation 1996 ; 93 : 1206 – 13 . Google Scholar CrossRef Search ADS PubMed 19 Fujimoto JG , Brezinski ME , Tearney GJ , Boppart SA , Bouma B , Hee MR et al. Optical biopsy and imaging using optical coherence tomography . Nat Med 1995 ; 1 : 970 – 2 . Google Scholar CrossRef Search ADS PubMed 20 Jang IK , Bouma BE , Kang DH , Park SJ , Park SW , Seung KB et al. Visualization of coronary atherosclerotic plaques in patients using optical coherence tomography: comparison with intravascular ultrasound . J Am Coll Cardiol 2002 ; 39 : 604 – 9 . Google Scholar CrossRef Search ADS PubMed 21 Gnanadesigan M , van Soest G , White S , Scoltock S , Ughi GJ , Baumbach A et al. Effect of temperature and fixation on the optical properties of atherosclerotic tissue: a validation study of an ex-vivo whole heart cadaveric model . Biomed Opt Express 2014 ; 5 : 1038 – 49 . Google Scholar CrossRef Search ADS PubMed Published on behalf of the European Society of Cardiology. All rights reserved. © The Author 2017. For permissions, please email: email@example.com.
European Heart Journal – Cardiovascular Imaging – Oxford University Press
Published: Nov 27, 2017
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera