Search

The adoption of emerging imaging technologies in the medical community is often hampered if they provide a new unfamiliar contrast that requires experience to be interpreted. Here, in order to facilitate such integration, we developed two complementary machine learning approaches, respectively based on feature engineering and on convolutional neural networks (CNN), to perform automatic diagnosis of breast biopsies using dynamic full field optical coherence tomography (D-FF-OCT) microscopy. This new technique provides fast, high resolution images of biopsies with a contrast similar to H&E histology, but without any tissue preparation and alteration. We conducted a pilot study on 51 breast biopsies, and more than 1,000 individual images, and performed standard histology to obtain each biopsy diagnosis. Using our automatic diagnosis algorithms, we obtained an accuracy above 88% at the image level, and above 96% at the biopsy level. Finally, we proposed different strategies to narrow down the spatial scale of the automatic segmentation in order to be able to draw the tumor margins by drawing attention maps with the CNN approach, or by performing high resolution precise annotation of the datasets. Altogether, these results demonstrate the high potential of D-FF-OCT coupled to machine learning to provide a rapid, automatic, and accurate histopathology diagnosis.

Dynamic full-field optical coherence microscopy (DFFOCM) was used to characterize the intracellular dynamic activities and cytoskeleton of HeLa cells in different viability states. HeLa cell samples were continuously monitored for 24 hours and compared with histological examination to confirm the cell viability states. The averaged mean frequency and magnitude observed in healthy cells were 4.79±0.5 Hz and 2.44±1.06, respectively. In dead cells, the averaged mean frequency was shifted to 8.57±0.71 Hz, whereas the magnitude was significantly decreased to 0.53±0.25. This cell dynamic activity analysis using DFFOCM is expected to replace conventional time-consuming and biopsies-required histological or biochemical methods.

International audience; Full-field optical coherence tomography (FFOCT) is an imaging technique of biological tissue based on tissue light reflectance analysis. We evaluated the feasibility of imaging fresh digestive mucosal biopsies after a quick mounting procedure (5 min) using two distinct modalities of FFOCT. In static FFOCT mode, we gained high-resolution images of general gut tissue-specific architecture, such as oesophageal papillae, gastric pits, duodenal villi and colonic crypts. In dynamic FFOCT mode, we imaged individual epithelial cells of the mucosal lining with a cellular or subcellular resolution and identified cellular components of the lamina propria. FFOCT represents a promising dye-free imaging tool for on-site analysis of gut tissue remodelling.

In this project, we analyzed 30 healthy and tumorous breast samples using static and dynamic full field optical coherence tomography (FF-OCT). We developed an automatic analysis workflow to classify each sample and compared it to an independent standard histological diagnosis. We used a first machine-learning algorithm to obtain cell and fiber segmentation of FF-OCT images and applied a linear support vector machine (SVM) analysis to classify each sample. We could obtain 100% specificity and sensitivity compared to histology. The label-free and non-invasive combination of static and dynamic FF-OCT thus appears very promising to obtain an efficient diagnosis of tumoral samples.

Dynamic Full-Field Optical Coherence Tomography (D-FFOCT) is a high transverse resolution version of OCT that records signal at the output of the interferometer as a movie over a few seconds. Analyzing the temporal variation of the signals reveals intracellular contrast which gives D-FFOCT the capability of identifying dynamic metabolic changes. We use D-FFOCT to identify oxygen induced changes in cellular metabolism at ambient (22%) oxygenation and at hypoxia (1%). Signal strength, calculated as an integral of the raw FFT spectrum, is significantly higher for 1% samples compared to 22% indicating that D-FFOCT is sensitive to changes in cell metabolism.

Introduction Malgre les progres de l’imagerie vasculaire, l’analyse histologique de biopsie d’artere temporale (BAT) reste l’examen de reference pour le diagnostic d’arterite a cellules geantes (ACG) [1] . Cependant, la BAT presente un certain nombre de limites - manque de sensibilite (50-85 %), caractere invasif, ou encore cout en temps et en personnel - si bien que des methodes d’imagerie des arteres temporales se developpent pour montrer des signes de vascularite (echo-Doppler, IRM, TEP scanner) [2] . Couramment utilisee en ophtalmologie clinique pour l’imagerie non invasive de coupes retiniennes, la tomographie de coherence optique (abregee OCT pour optical coherence tomography) exploite l’interference lumineuse comme principale source d’information pour l’observation des milieux biologiques. Nous rapportons pour la premiere fois son utilisation dans un montage haute resolution - l’OCT plein champ [3] - pour le diagnostic d’ACG, en comparaison de l’histologie conventionnelle. Materiels et methodes Des coupes congelees de 16 BAT ont ete analysees en OCT plein champ (Light-CTScanner, LLTech SAS, Paris, France). Apres coloration HES, les BAT etaient classees en 3 groupes: BAT negative sans hyperplasie intimale (n = 9), BAT negative avec hyperplasie intimale (n = 3) et BAT positive (n = 4). Une double analyse qualitative et quantitative des images obtenues en histologie (NDP.view2), puis OCT plein champ (ImageJ), a ensuite permis de confronter les deux techniques pour la visualisation ex vivo des modifications structurales de l’architecture de la paroi arterielle au cours de l’ACG. Enfin, l’application d’un filtre de Gabor a ete utilisee dans le but de cartographier l’orientation structurale comparee des fibres au sein de la paroi arterielle et ainsi tester l’hypothese d’une possible analyse automatisee des images acquises avec le systeme d’OCT plein champ. Resultats Les donnees acquises en OCT plein champ ont permis de visualiser les structures anatomiques presentes au sein de la paroi arterielle avec une resolution spatiale autorisant la detection d’une cellule unique: individualisation des trois tuniques (intima, media, adventice), limitante elastique interne (LEI) et vasa vasorum. La symetrie circulaire, la LEI et l’architecture de la paroi vasculaire etaient parfaitement conservees pour les 12 BAT negatives. En analyse quantitative, il existait une excellente correlation entre histologie conventionnelle et OCT plein champ pour la mesure de l’epaisseur des differentes tuniques, ce qui a permis de distinguer sans difficultes les BAT negatives avec ou sans hyperplasie intimale. Appuyee par l’application du filtre de Gabor et l’analyse d’image automatisee, l’observation des 4 BAT positives en OCT plein champ a montre une perte de la symetrie circulaire, avec homogeneisation des contrastes entre media et intima. Ce remodelage structural s’associait a une densification du tissu conjonctif qui s’etendait dans les trois directions de l’espace et finissait par effacer les contours de la LEI. Discussion Cette etude de preuve de concept rapporte la premiere utilisation de l’OCT plein champ pour le diagnostic histopathologique d’ACG. Ainsi, en quelques minutes seulement et sans traitement prealable de l’echantillon apres BAT, cette technique interferometrique hautement resolue permet l’acquisition directe de sections optiques transverses ex vivo, puis l’identification automatisee d’anomalies structurales caracteristiques d’ACG. Il devient notamment aise, meme pour des cliniciens non-inities au montage d’OCT plein champ, de retrouver des lesions elementaires d’ACG - remodelage structural de la paroi arterielle avec hyperplasie intimale et rupture de LEI - tres proches de celles observees en histologie apres colorations standards. Conclusion L’imagerie optique d’artere temporale par OCT plein champ constitue une alternative de choix pour le diagnostic rapide et automatise d’ACG. Des travaux complementaires sont en cours pour confirmer le potentiel de cette technologie et son application future.

DFFOCT allows for the visualization and characterization of cellular dynamics, providing future direction for cell-targeted therapeutics. In this work, we analyze the potential of DFFOCT to increase understanding of dynamic cellular processes under varying conditions.

Full-Field Optical Coherence Tomography (FFOCT) is an effective technique for tissue anatomy imaging, allowing cancer detection through the observation of disorders in the tissue microarchitecture. Moreover, with the temporal analysis of the FFOCT signal variations, a functional dimension is added, allowing the distinction of cell types through their intracellular activity. This complementary imaging mode, called Dynamic Cell Imaging (DCI), has shown the ability to identify normal cells, cancer cells and immune cells in different types of tissue such as breast, liver or lung. On samples ranging from cell cultures to entire tissue resections, DCI signals are recorded and analyzed in order to characterize the involved endogenous biomarker at the subcellular scale. Longitudinal studies of these samples over a few hours are performed under different environment perturbations intended to modify the cell metabolism. The potential of bimodal FFOCT is evaluated through a clinical study organized at the department of breast surgery of Peking University People’s Hospital in Beijing. More than 200 breast samples and lymph nodes are included. A part of the images will be directly compared to histology to identify reading criteria and build a reference atlas. The other part will be read on a blind study to measure the ability of cancer detection with respect to histology. To take full advantage of the FFOCT and the DCI information richness for faster cancer assessment during surgeries, a computer-aided diagnostic system based on Machine Learning, and more specifically Deep Learning, is investigated.

Non-melanoma skin cancer is the most common cancer. On cosmetically sensitive areas, Mohs micrographic surgery is the standard of care; intra-operative margin assessment minimizes the surgical defect while helping to reduce the recurrence rate by a factor of 3 compared to surgical excision. The current Mohs technique relies on frozen section tissue slide preparation, which significantly lengthens operative time and requires on-site trained histotechnicians. Full-field micro-tomography is a novel optical imaging technique based on interferometry. It allows the extraction of a full-field optical coherence tomography (FFOCT) image, representative of the morphology of the tissue, and the dynamic cell information, representative of the intra-cellular metabolic activity. Both images are calculated at the micron-level in a few minutes and without tissue preparation. This multi-centric study aimed to evaluate these combined new imaging modalities for the analysis of skin cancer margins during Mohs surgery. Over 200 Mohs specimens were imaged in Drexel University College of Medicine, USA, and GHR Mulhouse Sud Alsace, France. An atlas was established of FFOCT images and corresponding histological slides to reveal FFOCT reading criteria of normal and cancerous structures. Dynamic cell information enhanced visualization of cancerous cells and surrounding immune cells, and yielded metabolic quantification of cancerous area versus benign areas. Finally, deep learning algorithms were developed for preliminary results for the automatic detection of cancerous tissue. Hybrid morphologic and metabolic micro-tomography techniques hold great potential for skin cancer margin assessment. They can potentially reduce recurrence rates and surgery times, optimize clinical workflow, and decrease healthcare costs.

Full Field OCT (FFOCT) is a shot noise limited interferometric microscopy technique that uses incoherent light and has proved to be an effective diagnostic tool in terms of sensitivity and specificity. We have used the FFOCT setup built by LLTech for the analysis of various cancerous tissues corresponding to the following organs: breast, skin, prostate, lungs, ENT, bladder, brain etc. The scores obtained were found in the range between 80 and 98%. To do better and to provide informations that the histology does not carry we have studied, using the same setup, the temporal dependence of our signals which we found to be related to the cellular metabolism. We have used the new high speed and high full well capacity of the Adimec camera to achieve a time analysis ranging between 2 and a few thousands of ms. We thus obtain a new contrast which constitutes a biomarker at the sub-cellular scale. We monitor the characteristic frequencies and amplitude of the signal and display them on the images of the tissues using a new processing code of the time series. This metabolic contrast also reveal the evolution of the activity of cancer cells under treatments such as chemotherapy. We will illustrate this new approach through examples of cancer tissues that are planned to be used as intraoperative tools.

International audience

Ultrasound image (left) and acousto-optic image (right) of a liver biopsy with tumors. Acousto-optic imaging exhibits tumors that are not detected through ultrasound. Biological tissues are very strong light-scattering media. As a consequence, current medical imaging devices do not allow deep optical imaging unless invasive techniques are used. Acousto-optic imaging is a light-ultrasound coupling technique that takes advantage of the ballistic propagation of ultrasound in biological tissues to access optical contrast with a millimeter resolution. We have developed a photorefractive-crystal-based system that performs self-adaptive wavefront holography and works within the optical therapeutic window. As it works at an appropriate wavelength range for biological tissues imaging, it was tested on ex vivo liver samples containing tumors as a pre-clinical study. Optical contrast was obtained even if acoustical one was not significant.

August Wilson's The Piano Lesson features a debate between an African American brother and sister over the ownership of a richly symbolic piano, a family heirloom that represents the Charles family's slave heritage and its endurance through Reconstruction. Ownership questions like the one presented in The Piano Lesson can usually be resolved in the courts, but Wilson's play suggests that the law might be unable to resolve property disputes so problematically entangled with the legacy of slavery. Wilson offers, instead, a non-legal resolution to the piano debate presented in his play.

In this paper, we present a CMOS image sensor architecture coupling a spatial light modulator to a photodiode, for medical imaging based on acousto-optical coherence tomography with a digital holographic detection scheme. Our architecture is able to measure an interference pattern between a scattered beam transmitted through a scattering media and a reference beam, on an array with 16 μm pixel pitch, at 4000 Hz, which is compliant with correlation time of breast tissues. In-pixel processing allows generating from the incident light, a signal to polarize an embedded light modulator used to control the phase of the reflected beam. This reflected beam can then be focused on a region of interest of a scattering media, for therapy. The stacking of a photosensitive element with a spatial light modulator on the same device brings a significant robustness over the state of the art techniques such as perfect optical matching and reduced time delay in controlling light.

Photorefractive Bi(12)SiO(20) single crystal is used for acousto-optic imaging in thick scattering media in the green part of the spectrum, in an adaptive speckle correlation configuration. Light fields at the output of the scattering sample exhibit typical speckle grains of 1 μm size within the volume of the nonlinear crystal. This heterogeneous illumination induces a complex refractive index structure without applying a reference beam on the crystal, leading to a self-referenced diffraction correlation scheme. We demonstrate that this simple and robust configuration is able to perform axially resolved ultrasound modulated optical tomography of thick scattering media with an improved optical etendue.

Acousto-optic imaging is based on light interaction with focused ultrasound in a scattering medium. Thanks to photorefractive holography combined with pulsed ultrasound, we perform a time-resolved detection of ultrasound-modulated photons in the therapeutic window (780 nm). A high-gain SPS:Te crystal is used for this purpose and enables us to image through large optical thickness (500 mean free paths). We are able to generate three-dimensional (3D) acousto-optic images by translating a multielement ultrasound probe in only one direction. A 3D absorbing object is imaged through a 3 cm thick phantom. (C) 2012 Optical Society of America

Full-field optical coherence tomography (FF-OCT) is a powerful tool for nondestructive assessment of biological tissue, i.e., for the structural examination of tissue in depth at a cellular resolution. Mostly known as a microscopy device for ex vivo analysis, FF-OCT has also been adapted to endoscopy setups since it shows good potential for in situ cancer diagnosis and biopsy guidance. Nevertheless, all the attempts to perform endoscopic FF-OCT imaging did not go beyond lab setups. We describe here, to the best of our knowledge, the first handheld FF-OCT endoscope based on a tandem interferometry assembly using incoherent illumination. A common-path passive imaging interferometer at the tip of an optical probe makes it robust and insensitive to environmental perturbations, and a low finesse Fabry-Perot processing interferometer guarantees a compact system. A good resolution (2.7 μm transverse and 6 μm axial) is maintained through the long distance, small diameter relay optics of the probe, and a good signal-to-noise ratio is achieved in a limited 100 ms acquisition time. High-resolution images and a movie of a rat brain slice have been recorded by moving the contact endoscope over the surface of the sample, allowing for tissue microscopic exploration at 20 m under the surface. These promising ex vivo results open new perspectives for in vivo imaging of biological tissue, in particular, in the field of cancer and surgical margin assessment.

International audience; Acousto-optical coherence tomography (AOCT) consists in using random phase jumps on ultrasound and light to achieve a millimeter resolution when imaging thick scattering media. We combined this technique with heterodyne off-axis digital holography. Two-dimensional images of absorbing objects embedded in scattering phantoms are obtained with a good signal-to-noise ratio. We study the impact of the phase modulation characteristics on the amplitude of the acousto-optic signal and on the contrast and apparent size of the absorbing inclusion.

Acousto‐optic tomography is a technique that couples ultrasounds and light in order to measure local optical properties through thick and highly scattering media, e.g. human breast tissues. Thanks to the acousto‐optic effect, we can get the optical contrast information given by light and get the spatial localization from the ultrasound longitudinal waves. The use of a powerful long pulse laser (1 ms, 200 mJ) helps to improve the sensitivity of the technique by raising the optical peak power and collect more ultrasound tagged photons. Moreover, its tunable wavelength around 780 nm is appropriate for biological imaging since it is where light has its maximum penetrating depth in tissues. Detection of acousto‐optic signals is done by off‐axis heterodyne digital holography on a high speed CMOS camera. This enables us to make a tunable spatio‐temporal filter with a high signal to noise ratio. Experimentally, this technique enables to image optical absorbers embedded within a thick scattering media (a few centi...