Three-Dimensional Skin CT Based on Intelligent Algorithm in the Analysis of Skin Lesion Sites Features in Children with Psoriasis.

This research was to explore the application value of three-dimensional computed tomography (CT) based on artificial intelligent algorithm in analyzing the characteristics of skin lesions in children with psoriasis. In this study, 15 children with psoriasis were selected as the observation group, and 15 children with other skin diseases were selected as the control group. The CT images were optimized, and the feature selection was carried out based on artificial intelligent algorithm. Firstly, the results were compared with the results of simple skin three-dimensional CT to determine the effectiveness. Then, the two groups of three-dimensional skin CT image features of skin psoriasis-like hyperplasia, Munro microabscess, dermal papillary vascular dilation, and squamous epithelium based on intelligent algorithms were compared. After comparison, the detection rate of psoriasis-like hyperplasia, Munro microabscess, dermal papillary vascular dilation, and squamous epithelium in the observation group was higher than that in the control group, with significant difference and statistical significance (P < 0.05). In addition, the sensitivity of psoriasis-like hyperplasia, Munro microabscess, dermal papilla vascular dilatation, and squamous epithelium in children with psoriasis was 80.0%, 86.7%, 80.0%, and 93.3%, respectively. The specificity of psoriasis-like hyperplasia, Munro microabscess, dermal papilla vascular dilatation, and squamous epithelium in children with psoriasis was 86.7%, 93.3%, 60.0%, and 73.3%, respectively. The results showed that Munro microabscess and psoriasis-like hyperplasia had high sensitivity and specificity in all diagnostic items, which could be used as important features of skin lesion sites in the diagnosis of psoriasis in children. The research provides a basis for the clinical diagnosis of psoriasis in children, which is worthy of clinical promotion.

Primary Ovarian Melanoma Arising From a Mature Teratoma With Melanoma In Situ Present in the Ciliated Columnar and Squamous Epithelium in a Patient With Synchronous Skin Basal Cell Carcinoma.

Primary ovarian melanoma arising from ovarian teratomas are rarely reported and difficult to accurately diagnose. Cases in the literature rely on a diagnosis of exclusion, and cases of primary ovarian melanoma with pathologic evidence of melanoma in situ are exceedingly rare. We report a case of a 66-yr-old female who presented to emergency department with abdominal pain and bloating. Computed tomography scan showed a 21 cm complex pelvic mass. An urgent laparoscopic bilateral salpingo-oophorectomy was performed. Pathologically the mass was identified as a mature teratoma. Within the cystic teratoma, there was an area showing a sheet arrangement of atypical cells. Those atypical cells were positive for Melan A, Sox10, HMB45, and c-KIT, and negative for PD-L1. Melanoma in situ was present in both the squamous and ciliated columnar epithelium. The melanoma was negative for PD-L1, and no BRAF (codon 600, exons 11, 14, and 15) or c-KIT (exons 2, 9, 10, 11, 13, 14, 15, 17, 18) mutations were identified, thus supporting the so-called triple negative malignant melanoma. A thorough dermatologic exam was conducted and only a 3 mm skin basal cell carcinoma was confirmed on biopsy. At 11 mo of follow-up, the patient is disease free and doing well and no metastatic melanoma has been identified. To the best of our knowledge, this is the first documented case of a primary ovarian melanoma arising in a mature teratoma with evidence of melanoma in situ present in both ciliated columnar and squamous epithelium in a patient with synchronous skin basal cell carcinoma. Our case is positive for c-KIT protein (CD117) by immunohistochemistry, but negative for KIT mutation. More case reports are needed to further characterize the disease.

Acute pancreatitis in intraductal papillary mucinous neoplasms correlates with pancreatic volume and epithelial subtypes.

BACKGROUND: Intraductal papillary mucinous neoplasm (IPMN) of the pancreas is associated with acute pancreatitis (AP) in some cases, however its causes have not been fully elucidated. We investigated the association of the incidence of AP with epithelial subtypes and pancreatic volume in IPMN. METHODS: This retrospective study included 182 consecutive surgically resected IPMN patients between January 2000 and December 2018. The relationship between the incidence of AP and epithelial subtypes of IPMN and pancreatic volume was investigated. Epithelial subtypes of IPMN were classified into gastric (G type: N = 116), intestinal (I type: N = 49), pancreatobiliary (PB type: N = 14), and oncocytic types (O type: N = 3). Pancreatic volume of the contrast-enhanced computed tomography scan was measured using Ziostation2 software. Histological pancreatic parenchymal atrophy was also evaluated. RESULTS: AP occurred more frequently in I-types (I-type vs. G-type, 22.4% [11/49] vs 3.4% [4/116], P = 0.003) and PB-types (PB type vs. G-type, 35.7% [5/14] vs. 3.4% [4/116], P = 0.007) in comparison with G-types, which constituted the majority of the resected IPMNs. AP occurred more frequently in I-type patients with high pancreatic volumes (I-type with high pancreatic volume vs. I-type with low pancreatic volume, 37.0% [10/27] vs. 4.7% [1/21], P = 0.02). However, histological atrophy did not show an additional influence on the association between the incidence of AP and epithelial subtypes. The elevation of serum pancreatic enzymes was not significantly related to epithelial subtypes. CONCLUSION: Epithelial subtypes and the degree of pancreatic volume may be closely associated with the incidence of AP in IPMN.

Unraveling spatial cellular pattern by computational tissue shuffling.

Cell biology relies largely on reproducible visual observations. Unlike cell culture, tissues are heterogeneous, making difficult the collection of biological replicates that would spotlight a precise location. In consequence, there is no standard approach for estimating the statistical significance of an observed pattern in a tissue sample. Here, we introduce SET (for Synthesis of Epithelial Tissue), a method that can accurately reconstruct the cell tessellation formed by an epithelium in a microscopy image as well as thousands of alternative synthetic tessellations made of the exact same cells. SET can build an accurate null distribution to statistically test if any local pattern is necessarily the result of a process, or if it could be explained by chance in the given context. We provide examples in various tissues where visible, and invisible, cell and subcellular patterns are unraveled in a statistically significant manner using a single image and without any parameter settings.

Label-free, High-Resolution Optical Metabolic Imaging of Human Cervical Precancers Reveals Potential for Intraepithelial Neoplasia Diagnosis.

While metabolic changes are considered a cancer hallmark, their assessment has not been incorporated in the detection of early or precancers, when treatment is most effective. Here, we demonstrate that metabolic changes are detected in freshly excised human cervical precancerous tissues using label-free, non-destructive imaging of the entire epithelium. The images rely on two-photon excited fluorescence from two metabolic co-enzymes, NAD(P)H and FAD, and have micron-level resolution, enabling sensitive assessments of the redox ratio and mitochondrial fragmentation, which yield metrics of metabolic function and heterogeneity. Simultaneous characterization of morphological features, such as the depth-dependent variation of the nuclear:cytoplasmic ratio, is demonstrated. Multi-parametric analysis combining several metabolic metrics with morphological ones enhances significantly the diagnostic accuracy of identifying high-grade squamous intraepithelial lesions. Our results motivate the translation of such functional metabolic imaging to in vivo studies, which may enable improved identification of cervical lesions, and other precancers, at the bedside.

A rare case of BRAF V600E-mutated epithelioid glioblastoma with a sarcomatous component.

Epithelioid glioblastoma is a rare subtype of glioblastoma, but the coexistence of a sarcomatous component is even rarer. An 80-year-old woman was admitted to our hospital with somnolence. Magnetic resonance imaging revealed a cystic lesion with a solid component in the left temporal-parietal lobe. Histopathological examination of the resected tumor revealed three components; namely, typical glioblastoma, sarcomatous and epithelioid components at a ratio of about 5:3:2. All components were immunohistochemically positive for vimentin and mutated BRAF (V600E) and showed focal expression of glial fibrillary acidic protein and cytokeratin AE1/AE3, but they were negative for isocitrate dehydrogenase 1. Genetic analysis revealed that both the sarcomatous and epithelioid components harbored BRAF T1799A (V600E) mutation and homozygous deletion of cyclin-dependent kinase inhibitor 2A/B. We diagnosed this tumor as epithelial glioblastoma with a sarcomatous component. Our results indicate that even when the epithelial component is not dominant, immunohistochemical and genetic investigation of BRAF mutations is useful for the diagnosis of glioblastoma subtypes. In particular, although the prognosis of epithelial glioblastoma is poor, potentially effective targeted therapies for BRAF V600E-mutated tumors are available.

Prediction of the depth of invasion in superficial pharyngeal cancer: Microvessel morphological evaluation with narrowband imaging.

BACKGROUND: Magnifying endoscopy with narrowband imaging (ME-NBI) is useful in predicting the invasion depth by examining the microvascular status of tumor surfaces. This retrospective study aimed to determine its efficacy in pharyngeal cancer. METHODS: Between April 2016 and March 2018, 59 lesions from 46 patients who underwent transoral resection were retrospectively analyzed. Using ME-NBI, microvascular status was classified into B1, B2, or B3, based on the classification of the Japan Esophageal Society. RESULTS: A significant correlation was observed between microvascular status and invasion depth (P = .011). Mean thickness of lesions with B1, B2, and B3 vessels were 563, 1364, and 2825 µm, respectively (P = .006). In previously treated lesions, a significant correlation was observed between microvascular status and invasion depth (P = .012). CONCLUSIONS: ME-NBI is useful in predicting the invasion depth and thickness of pharyngeal tumors, even in patients with previously treated lesions.

Fast determination of coarse-grained cell anisotropy and size in epithelial tissue images using Fourier transform.

Mechanical strain and stress play a major role in biological processes such as wound healing or morphogenesis. To assess this role quantitatively, fixed or live images of tissues are acquired at a cellular precision in large fields of views. To exploit these data, large numbers of cells have to be analyzed to extract cell shape anisotropy and cell size. Most frequently, this is performed through detailed individual cell contour determination, using so-called segmentation computer programs, complemented if necessary by manual detection and error corrections. However, a coarse-grained and faster technique can be recommended in at least three situations: first, when detailed information on individual cell contours is not required; for instance, in studies which require only coarse-grained average information on cell anisotropy. Second, as an exploratory step to determine whether full segmentation can be potentially useful. Third, when segmentation is too difficult, for instance due to poor image quality or too large a cell number. We developed a user-friendly, Fourier-transform-based image analysis pipeline. It is fast (typically 10^{4} cells per minute with a current laptop computer) and suitable for time, space, or ensemble averages. We validate it on one set of artificial images and on two sets of fully segmented images, one from a Drosophila pupa and the other from a chicken embryo; the pipeline results are robust. Perspectives include in vitro tissues, nonbiological cellular patterns such as foams and xyz stacks.

Contrast of nuclei in stratified squamous epithelium in optical coherence tomography images at 800 nm.

Imaging nuclei of keratinocytes in the stratified squamous epithelium has been a subject of intense research since nucleus associated cellular atypia is the key criteria for the screening and diagnosis of epithelial cancers and their precursors. However, keratinocyte nuclei have been reported to be either low scattering or high scattering, so that these inconsistent reports might have led to misinterpretations of optical images, and more importantly, hindered the establishment of optical diagnostic criteria. We disclose that they are generally low scattering in the core using Micro-optical coherence tomography (µOCT) of 1.28-µm axial resolution in vivo; those previously reported "high scattering" or "bright" signals from nuclei are likely from the nucleocytoplasmic boundary, and the low-scattering nuclear cores were missed possibly due to insufficient axial resolutions (~4µm). It is further demonstrated that the high scattering signals may be associated with flattening of nuclei and cytoplasmic glycogen accumulation, which are valuable cytologic hallmarks of cell maturation.

Method for diagnosing neoplastic lesions by quantitative fluorescence value.

Fluorescence visualization devices (FVs) are useful for detecting malignant lesions because of their simple and noninvasive application. However, their quantitative application has been challenging. This study aimed to quantitatively and statistically evaluate the change in fluorescence intensity (FI) during the progression from normal epithelium to squamous cell carcinoma using a reproducible animal tongue carcinogenesis model. To establish this model, rats were treated with 50 ppm 4-Nitroquinoline 1-oxide (4NQO) in their drinking water for 10, 15, and 20 weeks. After 4NQO administration, each rat tongue was evaluated by gross observation, histology, and FI measurements. Fluorescence images were captured by FV, and ImageJ was used to measure FI, which was analyzed quantitatively and statistically. The establishment of a reproducible tumor progression model was confirmed, showing precancerous lesions (low-grade dysplasia [LGD]), early cancers (high-grade dysplasia/carcinoma in situ [HGD/CIS]), and advanced cancers (Cancer). This carcinogenesis model was quantitatively evaluated by FI. The FI of LGD stage was 54.6, which was highest intensity of all groups. Subsequently, the HGD/CIS and Cancer stages showed decreased FI (HGD/CIS: 46.1, Cancer: 49.1) and manifested as dark spots. This result indicates that FI had more variation and a wider range with increasing tumor progression. We demonstrated that FI migration and an uneven distribution are consistent with tumor progression. Since each step of tumor progression occurs reproducibly in this animal model, statistical evaluation was possible. In addition, tumor progression can be monitored by this new FI analysis method in humans.

A dense multi-path decoder for tissue segmentation in histopathology images.

BACKGROUND AND OBJECTIVE: Segmenting different tissue components in histopathological images is of great importance for analyzing tissues and tumor environments. In recent years, an encoder-decoder family of convolutional neural networks has increasingly adopted to develop automated segmentation tools. While an encoder has been the main focus of most investigations, the role of a decoder so far has not been well studied and understood. Herein, we proposed an improved design of a decoder for the segmentation of epithelium and stroma components in histopathology images. METHODS: The proposed decoder is built upon a multi-path layout and dense shortcut connections between layers to maximize the learning and inference capability. Equipped with the proposed decoder, neural networks are built using three types of encoders (VGG, ResNet and preactived ResNet). To assess the proposed method, breast and prostate tissue datasets are utilized, including 108 and 52 hematoxylin and eosin (H&E) breast tissues images and 224 H&E prostate tissue images. RESULTS: Combining the pre-activated ResNet encoder and the proposed decoder, we achieved a pixel wise accuracy (ACC) of 0.9122, a rand index (RAND) score of 0.8398, an area under receiver operating characteristic curve (AUC) of 0.9716, Dice coefficient for stroma (DICE_STR) of 0.9092 and Dice coefficient for epithelium (DICE_EPI) of 0.9150 on the breast tissue dataset. The same network obtained 0.9074 ACC, 0.8320 Rand index, 0.9719 AUC, 0.9021 DICE_EPI and 0.9121 DICE_STR on the prostate dataset. CONCLUSIONS: In general, the experimental results confirmed that the proposed network is superior to the networks combined with the conventional decoder. Therefore, the proposed decoder could aid in improving tissue analysis in histopathology images.

WASH phosphorylation balances endosomal versus cortical actin network integrities during epithelial morphogenesis.

Filamentous actin (F-actin) networks facilitate key processes like cell shape control, division, polarization and motility. The dynamic coordination of F-actin networks and its impact on cellular activities are poorly understood. We report an antagonistic relationship between endosomal F-actin assembly and cortical actin bundle integrity during Drosophila airway maturation. Double mutants lacking receptor tyrosine phosphatases (PTP) Ptp10D and Ptp4E, clear luminal proteins and disassemble apical actin bundles prematurely. These defects are counterbalanced by reduction of endosomal trafficking and by mutations affecting the tyrosine kinase Btk29A, and the actin nucleation factor WASH. Btk29A forms protein complexes with Ptp10D and WASH, and Btk29A phosphorylates WASH. This phosphorylation activates endosomal WASH function in flies and mice. In contrast, a phospho-mimetic WASH variant induces endosomal actin accumulation, premature luminal endocytosis and cortical F-actin disassembly. We conclude that PTPs and Btk29A regulate WASH activity to balance the endosomal and cortical F-actin networks during epithelial tube maturation.

Cytoplasmic Fibrillar Aggregates in Gallbladder Epithelium Are a Frequent Mimic of Cystoisospora in Pediatric Cholecystectomy Specimens.

CONTEXT.­: Cystoisospora belli is an intracellular parasite associated with gastrointestinal disease in immunocompromised hosts. Although infection has been classically associated with intestinal disease, studies have identified Cystoisospora in the gallbladder of immunocompetent patients based on hematoxylin-eosin morphology. Recently, the identity of this histologic finding as Cystoisospora has been questioned based on negative results of nucleic acid studies. OBJECTIVE.­: To determine the prevalence of this histologic feature in pediatric patients, we retrospectively reviewed all cholecystectomy specimens from a pediatric hospital during a 24-month period. DESIGN.­: In 180 cholecystectomy specimens, we identified 11 cases (6.1%) with classical histologic features previously described to represent Cystoisospora organisms. To further investigate these structures, we retrieved tissue from paraffin-embedded blocks and performed electron microscopy. RESULTS.­: Ultrastructural examination identified ovoid perinuclear cytoplasmic structures composed of dense fibrillar aggregates rather than organisms. Patients with positive cases were similar in age to controls (positive cases: mean patient age 13.4 years [range, 2-23 years]; negative cases: mean patient age 14.7 years [range, 12 weeks-31 years]; P = .35). There was no significant association of this finding with cholelithiasis (54.5% versus 65.1%, P = .52), cholesterolosis (0% versus 22.5%, P = .12), acute cholecystitis (9.1% versus 10.1%, P > .99), or chronic cholecystitis (45.5% versus 66.3%, P = .20). CONCLUSIONS.­: To our knowledge, this is the first positive identification of these structures as cytoplasmic fibrillar aggregates rather than parasitic inclusions by ultrastructural examination, and the first study of this histologic finding in pediatric cholecystectomies.

Three-dimensional architecture of common benign and precancerous prostate epithelial lesions.

AIMS: Many glandular lesions can mimic prostate cancer microscopically, including atrophic glands, adenosis and prostatic intraepithelial neoplasia. While the characteristic histopathological and immunohistochemical features of these lesions have been well established, little is known about their three-dimensional architecture. Our objective was to evaluate the three-dimensional organisation of common prostate epithelial lesions. METHODS AND RESULTS: 500 µm-thick punches (n = 42) were taken from radical prostatectomy specimens, and stained with antibodies targeting keratin 8-18 and keratin 5 for identification of luminal and basal cells, respectively. Tissue samples were optically cleared in benzyl alcohol:benzyl benzoate and imaged using a confocal laser scanning microscope. The three-dimensional architecture of peripheral and transition zone glands was acinar, composed of interconnecting and blind-ending saccular tubules. In simple atrophy, partial atrophy and post-atrophic hyperplasia, the acinar structure was attenuated with branching blind-ending tubules from parental tubular structures. Three-dimensional imaging revealed a novel variant of prostate atrophy characterised by large Golgi-like atrophic spaces parallel to the prostate surface, which were represented by thin, elongated tubular structures on haematoxylin and eosin (H&E) slides. Conversely, adenosis lacked acinar organisation, so that it closely mimicked low-grade prostate cancer. High-grade prostatic intraepithelial neoplasia displayed prominent papillary intraluminal protrusions but retained an acinar organisation, whereas intraductal carcinoma predominantly consisted of cribriform proliferations with either spheroid, ellipsoid or complex interconnecting lumens. CONCLUSIONS: While various prostate epithelial lesions might mimic malignancy on H&E slides, their three-dimensional architecture is acinar and clearly different from the tubular structure of prostate cancer, with adenosis as an exception.

New Insight into the Development of the Respiratory Acini in Rabbits: Morphological, Electron Microscopic Studies, and TUNEL Assay.

This study investigated the histomorphological features of developing rabbit respiratory acini during the postnatal period. On the 1st day of postnatal life, the epithelium of terminal bronchiole consisted of clear cells which intercalated between few ciliated and abundant non-ciliated (Clara) cells. At this age, the rabbit lung was in the alveolar stage. The terminal bronchioles branched into several alveolar ducts, which opened into atria that communicated to alveolar sacs. All primary and secondary inter-alveolar septa were thick and showed a double-capillary network (immature septa). The primitive alveoli were lined largely by type-I pneumocytes and mature type-II pneumocytes. The type-I pneumocytes displayed an intimate contact with the endothelial cells of the blood capillaries forming the blood-air barrier (0.90 ± 0.03 µm in thickness). On the 3rd day, we observed intense septation and massive formation of new secondary septa giving the alveolar sac a crenate appearance. The mean thickness of the air-blood barrier decreased to reach 0.78 ± 0.14 µm. On the 7th day, the terminal bronchiole epithelium consisted of ciliated and non-ciliated cells. The non-ciliated cells could be identified as Clara cells and serous cells. New secondary septa were formed, meanwhile the inter-alveolar septa become much thinner and the air-blood barrier thickness was 0.66 ± 0.03 µm. On the 14th day, the terminal bronchiole expanded markedly and the pulmonary alveoli were thin-walled. Inter-alveolar septa become much thinner and single capillary layers were observed. In the 1st month, the secondary septa increased in length forming mature cup-shaped alveoli. In the 2nd month, the lung tissue grew massively to involve the terminal respiratory unit. In the 3rd month, the pulmonary parenchyma appeared morphologically mature. All inter-alveolar septa showed a single-capillary layer, and primordia of new septa were also observed. The thickness of the air-blood barrier was much thinner; 0.56 ± 0.16 µm. TUNEL assay after birth revealed that the apoptotic cells were abundant and distributed in the epithelium lining of the pulmonary alveoli and the interstitium of the thick interalveolar septa. On the 7th day, and onward, the incidence of apoptotic cells decreased markedly. This study concluded that the lung development included two phases: the first phase (from birth to the 14th days) corresponds to the period of bulk alveolarization and microvascular maturation. The second phase (from the 14th days to the full maturity) corresponds to the lung growth and late alveolarization.

Tumour parcellation and quantification (TuPaQ): a tool for refining biomarker analysis through rapid and automated segmentation of tumour epithelium.

BACKGROUND AND AIMS: Immunohistochemistry (IHC) is an essential component of biomarker research in cancer. Automated biomarker quantification is hampered by the failure of computational algorithms to discriminate 'negative' tumour cells from 'negative' stromal cells. We sought to develop an algorithm for segmentation of tumour epithelium in colorectal cancer (CRC), irrespective of the biomarker expression in the cells. METHODS AND RESULTS: We developed tumour parcellation and quantification (TuPaQ) to segment tumour epithelium and parcellate sections into 'epithelium' and 'non-epithelium'. TuPaQ comprises image pre-processing, extraction of regions of interest (ROIs) and quantification of tumour epithelium (total area occupied by epithelium and number of nuclei in the occupied area). A total of 286 TMA cores from CRC were manually annotated and analysed using the commercial halo software to provide ground truth. The performance of TuPaQ was evaluated against the ground truth using a variety of metrics. The image size of each core was 7000 × 7000 pixels and each core was analysed in a matter of seconds. Pixel × pixel analysis showed a sensitivity of 84% and specificity of 95% in detecting epithelium. The mean tumour area obtained by TuPaQ was very close to the area quantified after manual annotation (r = 0.956, P < 0.001). Moreover, quantification of tumour nuclei by TuPaQ correlated very strongly with that of halo (r = 0.891, P < 0.001). CONCLUSION: TuPaQ is a very rapid and accurate method of separating the epithelial and stromal compartments of colorectal tumours. This will allow more accurate and objective analysis of immunohistochemistry.

Outcome of endoscopic mucosal resection in Barrett's esophagus determined by systematic quantification of epithelial glands using volumetric laser endomicroscopy.

BACKGROUND: Dysplastic Barrett's esophagus (BE) lesions ≤2 cm in size can be targeted for en-bloc endoscopic mucosal resection (EMR). White-light endoscopy can underestimate the size of a lesion, limiting complete resection. Volumetric laser endomicroscopy (VLE) provides high-resolution cross-sectional imaging of BE. Epithelial glands are a VLE feature associated with BE dysplasia. We study the association between VLE gland quantification and outcome of resection. METHODS: EMR specimens of BE lesions targeted for en-bloc resection were imaged with VLE using an established protocol. Manual and automated quantification of epithelial glands was performed blinded to resection outcome. The presence of epithelial glands at the resection margins was recorded. Histologic en-bloc (R0) resection of the targeted lesion was defined by the absence and incomplete (R1) resection by the presence of dysplasia/neoplasia at specimen margins. RESULTS: Thirty-seven EMRs with a mean (standard deviation) size of 1.04 (0.37) cm were imaged with VLE. The highest grade of dysplasia found was low-grade dysplasia (n = 12), high-grade dysplasia (n = 19), and intramucosal cancer (n = 6). The en-bloc resection rate was 37.8% (R0, n = 14; R1, n = 23). The mean (standard deviation) number of epithelial glands quantified with VLE was 13.0 (6.7) and 28.8 (23.9) for R0 and R1 specimens, respectively, with a significant mean difference of 15.8 glands (95% confidence interval, 2-29; P = .02). The presence of glands at the specimen margin was associated with incomplete resection (P < .001). CONCLUSION: Systematic quantification of BE epithelial glands using VLE can determine the outcome of endoscopic resection. VLE may have a potential role in assessment of lesion margins.

Light scattering measured with spatial frequency domain imaging can predict stromal versus epithelial proportions in surgically resected breast tissue.

This study aims to determine if light scatter parameters measured with spatial frequency domain imaging (SFDI) can accurately predict stromal, epithelial, and adipose fractions in freshly resected, unstained human breast specimens. An explicit model was developed to predict stromal, epithelial, and adipose fractions as a function of light scattering parameters, which was validated against a quantitative analysis of digitized histology slides for N = 31 specimens using leave-one-out cross-fold validation. Specimen mean stromal, epithelial, and adipose volume fractions predicted from light scattering parameters strongly correlated with those calculated from digitized histology slides (r = 0.90, 0.77, and 0.91, respectively, p-value <1 × 10 - 6). Additionally, the ratio of predicted epithelium to stroma classified malignant specimens with a sensitivity and specificity of 90% and 81%, respectively, and also classified all pixels in malignant lesions with 63% and 79%, at a threshold of 1. All specimens and pixels were classified as malignant, benign, or fat with 84% and 75% accuracy, respectively. These findings demonstrate how light scattering parameters acquired with SFDI can be used to accurately predict and spatially map stromal, epithelial, and adipose proportions in fresh unstained, human breast tissue, and suggest that these estimations could provide diagnostic value.

Radio-pathomic Maps of Epithelium and Lumen Density Predict the Location of High-Grade Prostate Cancer.

PURPOSE: This study aims to combine multiparametric magnetic resonance imaging (MRI) and digitized pathology with machine learning to generate predictive maps of histologic features for prostate cancer localization. METHODS AND MATERIALS: Thirty-nine patients underwent MRI prior to prostatectomy. After surgery, tissue was sliced according to MRI orientation using patient-specific 3-dimensionally printed slicing jigs. Whole-mount sections were annotated by our pathologist and digitally contoured to differentiate the lumen and epithelium. Slides were co-registered to the T2-weighted MRI scan. A learning curve was generated to determine the number of patients required for a stable machine-learning model. Patients were randomly stratified into 2 training sets and 1 test set. Two partial least-squares regression models were trained, each capable of predicting lumen and epithelium density. Predicted density values were calculated for each patient in the test dataset, mapped into the MRI space, and compared between regions confirmed as high-grade prostate cancer. RESULTS: The learning-curve analysis showed that a stable fit was achieved with data from 10 patients. Maps indicated that regions of increased epithelium and decreased lumen density, generated from each independent model, corresponded with pathologist-annotated regions of high-grade cancer. CONCLUSIONS: We present a radio-pathomic approach to mapping prostate cancer. We find that the maps are useful for highlighting high-grade tumors. This technique may be relevant for dose-painting strategies in prostate radiation therapy.