Establishment of artificial intelligence model for precise histological subtyping of lung adenocarcinoma and its application to quantitative and spatial analysis.

BACKGROUND: The histological subtype of lung adenocarcinoma is a major prognostic factor. We developed a new artificial intelligence model to classify lung adenocarcinoma images into seven histological subtypes and adopted the model for whole-slide images to investigate the relationship between the distribution of histological subtypes and clinicopathological factors. METHODS: Using histological subtype images, which are typical for pathologists, we trained and validated an artificial intelligence model. Then, the model was applied to whole-slide images of resected lung adenocarcinoma specimens from 147 cases. RESULT: The model achieved an accuracy of 99.7% in training sets and 90.4% in validation sets consisting of typical tiles of histological subtyping for pathologists. When the model was applied to whole-slide images, the predominant subtype according to the artificial intelligence model classification matched that determined by pathologists in 75.5% of cases. The predominant subtype and tumor grade (using the WHO fourth and fifth classifications) determined by the artificial intelligence model resulted in similar recurrence-free survival curves to those determined by pathologists. Furthermore, we stratified the recurrence-free survival curves for patients with different proportions of high-grade components (solid, micropapillary and cribriform) according to the physical distribution of the high-grade component. The results suggested that tumors with centrally located high-grade components had a higher malignant potential (P < 0.001 for 5-20% high-grade component). CONCLUSION: The new artificial intelligence model for histological subtyping of lung adenocarcinoma achieved high accuracy, and subtype quantification and subtype distribution analyses could be achieved. Artificial intelligence model therefore has potential for clinical application for both quantification and spatial analysis.

Three Distinct Stroma Types in Human Pancreatic Cancer Identified by Image Analysis of Fibroblast Subpopulations and Collagen.

PURPOSE: Cancer-associated fibroblasts have emerged to be highly heterogenous and can play multifaceted roles in dictating pancreatic ductal adenocarcinoma (PDAC) progression, immunosuppression, and therapeutic response, highlighting the need for a deeper understanding of stromal heterogeneity between patients and even within a single tumor. We hypothesized that image analysis of fibroblast subpopulations and collagen in PDAC tissues might guide stroma-based patient stratification to predict clinical outcomes and tumor characteristics. EXPERIMENTAL DESIGN: A novel multiplex IHC-based image analysis system was established to digitally differentiate fibroblast subpopulations. Using whole-tissue slides from 215 treatment-naïve PDACs, we performed concurrent quantification of principal fibroblast subpopulations and collagen and defined three stroma types: collagen-rich stroma, fibroblast activation protein α (FAP)-dominant fibroblast-rich stroma, and α smooth muscle actin (ACTA2)-dominant fibroblast-rich stroma. These stroma types were assessed for the associations with cancer-specific survival by multivariable Cox regression analyses and with clinicopathologic factors, including CD8+ cell density. RESULTS: FAP-dominant fibroblasts and ACTA2-dominant fibroblasts represented the principal distinct fibroblast subpopulations in tumor stroma. Stroma types were associated with patient survival, SMAD4 status, and transcriptome signatures. Compared with FAP-dominant fibroblast-rich stroma, collagen-rich stroma correlated with prolonged survival [HR, 0.57; 95% confidence interval (CI), 0.33-0.99], while ACTA2-dominant fibroblast-rich stroma exhibited poorer prognosis (HR, 1.65; 95% CI, 1.06-2.58). FAP-dominant fibroblast-rich stroma was additionally characterized by restricted CD8+ cell infiltrates and intense neutrophil infiltration. CONCLUSIONS: This study identified three distinct stroma types differentially associated with survival, immunity, and molecular features, thereby underscoring the importance of stromal heterogeneity in subtyping pancreatic cancers and supporting the development of antistromal therapies.

Pre-hepatectomy type IV collagen 7S predicts post-hepatectomy liver failure and recovery.

BACKGROUND: Liver resection is an effective treatment for benign and malignant liver tumors. However, a method for preoperative evaluation of hepatic reserve has not yet been established. Previously reported assessments of preoperative hepatic reserve focused only on liver failure in the early postoperative period and did not consider the long-term recovery of hepatic reserve. When determining eligibility for hepatectomy, the underlying pathophysiology needs to be considered to determine if the functional hepatic reserve can withstand both surgery and any postoperative therapy. AIM: To identify pre-hepatectomy factors associated with both early postoperative liver failure and long-term postoperative liver function recovery. METHODS: This study was a retrospective cohort study. We retrospectively investigated 215 patients who underwent hepatectomy at our hospital between May 2013 and December 2016. Early post-hepatectomy liver failure (PHLF) was defined using the International Study Group of Liver Surgery's definition of PHLF. Long-term postoperative recovery of liver function was defined as the time taken for serum total bilirubin and albumin levels to return to levels of < 2 mg/dL and > 2.8 g/dL, respectively, and the time taken for Child-Pugh score to return to Child-Pugh class A. RESULTS: Preoperative type IV collagen 7S was identified as a significant independent factor associated with both PHLF and postoperative long-term recovery of liver function. Further analysis revealed that the time taken for the recovery of Child-Pugh scores and serum total bilirubin and albumin levels was significantly shorter in patients with type IV collagen 7S ≤ 6 ng/mL than in those with type IV collagen 7S > 6 ng/mL. In additional analyses, similar results were observed in patients without chronic viral hepatitis associated with fibrosis. CONCLUSION: Preoperative type IV collagen 7S is a preoperative predictor of PHLF and long-term postoperative liver function recovery. It can also be used in patients without chronic hepatitis virus.

Quantification of intratumoral collagen and elastin fibers within hepatocellular carcinoma tissues finds correlations with clinico-patho-radiological features.

AIM: Emerging evidence suggests a promising role for tumor stromal factors in characterizing patients with various types of malignancies, including hepatocellular carcinoma (HCC). We quantified the amount of collagen and elastin fibers in HCC samples with the aim of clarifying the clinico-patho-radiological significance of fiber deposition in HCC. METHODS: We computed the amount of collagen and elastin fibers using digital image analysis of whole-slide images of Elastica van Gieson-stained tissues from 156 surgically resected HCCs. Furthermore, we assessed the correlations between the fiber content of HCC samples and clinical, pathological, and radiological features, including immunohistochemistry-based molecular subtypes and immunosubtypes. RESULTS: The intratumoral area ratio of collagen in HCC tissues (median 3.4%, range 0.1-22.2%) was more than threefold that of elastin (median 0.9%, range 0.1-9.0%); there was a strong positive correlation between the amounts of collagen and elastin. Higher levels of combined collagen and elastin were significantly associated with the confluent multinodular macroscopic tumor type, the absence of a fibrous capsule, intratumoral steatosis, scirrhous tumor stroma, dense inflammatory-cell infiltrates, and the biliary/stem cell markers-positive HCC subtype. The associations of higher collagen levels with radiological findings, including heterogeneous enhancement and persistent enhancement on dynamic computed tomography, were significant. In contrast, the associations of radiological findings with elastin fibers were not significant. Intratumoral fibrous stroma in HCC comprised septum-like and perisinusoidal fibrosis; these two forms represented distinct distribution patterns of fibers and fibroblasts. CONCLUSION: Quantitative analysis suggested that stromal fiber-rich HCCs likely represent a distinct clinico-patho-radiological entity.

Maternal undernutrition aggravates renal tubular necrosis and interstitial fibrosis after unilateral ureteral obstruction in male rat offspring.

Maternal undernutrition is known to reduce glomerular number but it may also affect tubulointerstitium, capillary density, and response to oxidative stress. To investigate whether the latter elements are affected, we examined the response to unilateral ureteral obstruction (UUO), an established model of renal tubulointerstitial fibrosis, in the kidney of offspring from control and nutrient restricted rats. Six-week old male offspring from rats given food ad libitum (CON) and those subjected to 50% food restriction throughout pregnancy (NR) were subjected to UUO for 7 days. Body weight was significantly lower in NR. Systolic blood pressure and blood urea nitrogen increased similarly in CON and NR after UUO. Tubular necrosis in the obstructed kidney, on the other hand, was more extensive in NR. Also, the collagen area, a marker of fibrosis, of the obstructed kidney was significantly increased compared with the contralateral kidney only in NR. Capillary density was decreased similarly in the obstructed kidney of CON and NR compared with the contralateral kidney. Urine nitrate/nitrite, a marker of nitric oxide production, from the obstructed kidney was significantly increased in NR compared with CON. Nitrotyrosine, a marker of nitric oxide-mediated free radical injury, was increased in the obstructed kidney compared with the contralateral kidney in both CON and NR, but the extent was significantly greater in NR. In conclusion, more severe tubular necrosis and fibrosis after UUO was observed in NR, which was thought to be due to increased nitrosative stress.

Characterization of spatial distribution of tumor-infiltrating CD8+ T cells refines their prognostic utility for pancreatic cancer survival.

The abundance of cytotoxic T-cell infiltrates has important implications for patient outcome and therapeutic design for pancreatic ductal adenocarcinoma. However, intratumoral heterogeneity remains a challenge to understanding the complex immune microenvironment. We hypothesized that characterizing CD8+ cell distribution within pancreatic adenocarcinoma tissues might refine the prognostic value of tumor-infiltrating CD8+ lymphocytes. Using multiplex immunohistochemistry-based image analysis on whole-tissue sections of 214 pancreatic ductal adenocarcinomas, we measured CD8+ cell densities in the tumor center, the tumor margin, and the whole tumor, along with the proximity of CD8+ cells to carcinoma cells. Multivariable Cox regression analysis was performed to assess the associations of CD8+ cell densities with pancreatic cancer-specific survival, adjusting for clinicopathologic and immune-related features, including tumor expressions of TP53, SMAD4, and the programmed cell death 1 ligand 1 (CD274, PD-L1) and the extent of tertiary lymphoid structures. There was substantial heterogeneity in CD8+ cell density, with the mean density in the tumor center less than half that in the tumor margin. Tumor CD274 expression and extensive tertiary lymphoid structures were appeared to be associated with higher CD8+ cell density in the tumor margin (P = 0.037 and P = 0.005, respectively), but not with that in the tumor center (P > 0.50). The association of higher CD8+ cell density with prolonged survival was significant for the whole tumor (Ptrend = 0.009); however, the association was stronger for the tumor center (Ptrend = 0.002) and insignificant for the tumor margin (Ptrend = 0.07). Tumor cell-CD8+ cell distance correlated strongly with CD8+ cell density, whereas the density of CD8+ cells proximate to cancer cells exhibited no prognostic association. In conclusion, spatial computational analysis on pancreatic ductal adenocarcinoma reveals the prognostic validity of CD8+ cell density in the tumor center, where CD8+ cell infiltration is ununiformly restricted, likely suggesting pro-tumorigenic roles of the immunosuppressive tumor microenvironment of pancreatic cancer.

Non-invasive liver fibrosis assessment correlates with collagen and elastic fiber quantity in patients with hepatitis C virus infection.

AIM: Elastic fiber deposition is a cause of irreversibility of liver fibrosis. However, to date, its relevance to clinical features has not yet been clarified. This study aimed to clarify the correlation between non-invasive markers of fibrosis and fiber quantity, including elastic fiber, obtained from computational analysis. METHODS: This retrospective study included 270 patients evaluated by non-invasive liver fibrosis assessment prior to liver biopsy. Of these patients, 95 underwent magnetic resonance elastography (MRE) and 244 were assessed with Wisteria floribunda agglutinin-positive Mac-2 binding protein (WFA+ -M2BP). Using whole-slide imaging of Elastica van Gieson-stained liver biopsy sections, the quantity of collagen, elastin, and total fiber (elastin + collagen) was determined. RESULTS: The total fiber quantity showed significant linear correlation with fibrosis stage F0-F4. Collagen fiber quantity increased from stage F0 to F4, whereas elastic fiber quantity increased significantly only from stage F2 to F3. Spearman's rank correlation test revealed that non-invasive liver fibrosis assessment significantly correlates with each fiber quantity, including correlation between total fiber quantity and the Fibrosis-4 (FIB-4) index (r = 0.361, P < 0.001), WFA+ -M2BP values (r = 0.404, P < 0.001), and liver stiffness value by MRE (r = 0.615, P < 0.001). Receiver operating characteristic (ROC) curve analyses revealed that the area under ROC for predicting higher elastic fiber (>3.6%) is 0.731 by FIB-4 index, 0.716 by WFA+ -M2BP, and 0.822 by liver stiffness by MRE. CONCLUSION: Liver fibrosis correlates with fiber quantity through non-invasive assessment regardless of fiber type, including elastic fiber. Moreover, MRE is useful for predicting high amounts of elastic fiber.

Evaluation of Pancreatic Fibrosis With Acoustic Radiation Force Impulse Imaging and Automated Quantification of Pancreatic Tissue Components.

OBJECTIVES: The aim of this study was to determine whether computer-assisted digital analysis and acoustic radiation force impulse (ARFI) imaging were useful for assessing pancreatic fibrosis, and if ARFI imaging predicted postoperative pancreatic fistula (POPF). METHODS: Seventy-eight patients scheduled to undergo pancreatic resection were enrolled. Shear wave velocity (SWV) at the pancreatic neck was measured preoperatively using ARFI imaging. Pancreatic tissue components on a whole slide image were quantified using an automatic image processing software. The relationship between SWV, fibrotic tissue content, and POPF incidence and clinical severity was analyzed. RESULTS: The median collagen fiber, fatty tissue, and acinar cell contents were 11.6%, 8.5%, and 61.3%, respectively. Unlike fatty tissue, collagen fiber content and acinar cells were correlated with SWV (ρ = 0.440, P < 0.001 and ρ = -0.428, P < 0.001, respectively). Although collagen fiber content and SWV were associated with the overall incidence of POPF (P = 0.004 and 0.001, respectively), collagen fiber content and SWV had no statistical correlation with clinically relevant POPF (P = 0.268 and 0.052, respectively). CONCLUSIONS: We objectively quantified the pancreatic tissue components using an automatic image processing software. Shear wave velocity was significantly related to collagen fiber content and suggests that ARFI imaging can be useful for evaluating pancreatic fibrosis.

Quantitative assessment of liver fibrosis reveals a nonlinear association with fibrosis stage in nonalcoholic fatty liver disease.

Accurate staging of liver fibrosis is crucial to guide therapeutic decisions for patients with nonalcoholic fatty liver disease (NAFLD). Digital image analysis has emerged as a promising tool for quantitative assessment of fibrosis in chronic liver diseases. We sought to determine the relationship of histologic fibrosis stage with fiber amounts quantified in liver biopsy specimens for the better understanding of NAFLD progression. We measured area ratios of collagen and elastin fibers in Elastica van Gieson-stained biopsy tissues from 289 patients with NAFLD from four hospitals using an automated computational method and examined their correlations with Brunt's fibrosis stage. As a secondary analysis, we performed multivariable logistic regression analysis to assess the associations of the combined area ratios of collagen and elastin with noninvasive fibrosis markers. The combined fiber area ratios correlated strongly with Brunt's stage (Spearman correlation coefficient, 0.78; P < 0.0001), but this relationship was nonlinear (P = 0.007) with striking differences between stage 4 (median area ratios, 12.3%) and stages 0-3 (2.1%, 2.8%, 4.3%, and 4.8%, respectively). Elastin accumulation was common in areas of thick bridging fibrosis and thickened venous walls but not in areas of perisinusoidal fibrosis. The highest tertile of the combined fiber area ratios was associated with the fibrosis-4 index and serum type IV collagen 7s domain (7s collagen) levels, whereas the upper two tertiles of the fiber amounts significantly associated with body mass index, aspartate aminotransferase, and 7s collagen in the multivariable analysis. Conclusion: Quantitative fibrosis assessment reveals a nonlinear relationship between fibrosis stage and fiber amount, with a marked difference between stage 4 and stage 3 and much smaller differences among stages 0-3, suggesting a heterogeneity in disease severity within NAFLD-related cirrhosis. (Hepatology Communications 2018;2:58-68).

A case of Timothy syndrome with adrenal medullary dystrophy.

Timothy syndrome (TS) is a congenital long QT syndrome that is associated with syndactyly and mutations in CACNA1C, encoding an L-type voltage-dependent calcium channel, Cav1.2. Recently, TS has been associated with autism and other psychological disorders. This case indicated bradycardia by prenatal screening and was diagnosed as TS by the occurrence of syndactyly and QT prolongation at birth. Despite therapy with anti-arrhythmia reagents and a pacemaker, the patient died 2 months after birth and was autopsied. The heart showed mild dilation and mild hypertrophy with a focal disarray pattern, which may be inconsistent with typical cardiomyopathy. Unexpectedly, bilateral adrenal glands showed marked shrinkage and severe fibrosis of the medulla with a small number of single-strand DNA positive medullary cells and accumulation of hemosiderin-containing macrophages. This finding suggests that CACNA1C mutation may induce drop-out of medulla cells via apoptosis. This may be due to increased concentration of calcium ions consistent with Cav1.2 expression in adrenal glands as well as in the brain and the heart. This is the first report describing a systemic autopsy of TS with adrenal medullary dystrophy.

Correction: Elastin Fiber Accumulation in Liver Correlates with the Development of Hepatocellular Carcinoma.

[This corrects the article DOI: 10.1371/journal.pone.0154558.].

Automatic quantification of morphological features for hepatic trabeculae analysis in stained liver specimens.

This paper proposes a digital image analysis method to support quantitative pathology by automatically segmenting the hepatocyte structure and quantifying its morphological features. To structurally analyze histopathological hepatic images, we isolate the trabeculae by extracting the sinusoids, fat droplets, and stromata. We then measure the morphological features of the extracted trabeculae, divide the image into cords, and calculate the feature values of the local cords. We propose a method of calculating the nuclear-cytoplasmic ratio, nuclear density, and number of layers using the local cords. Furthermore, we evaluate the effectiveness of the proposed method using surgical specimens. The proposed method was found to be an effective method for the quantification of the Edmondson grade.

Elastin Fiber Accumulation in Liver Correlates with the Development of Hepatocellular Carcinoma.

BACKGROUND & AIMS: The fibrosis stage, which is evaluated by the distribution pattern of collagen fibers, is a major predictor for the development of hepatocellular carcinoma (HCC) for patients with hepatitis C. Meanwhile, the role of elastin fibers has not yet been elucidated. The present study was conducted to determine the significance of quantifying both collagen and elastin fibers. METHODS: We enrolled 189 consecutive patients with hepatitis C and advanced fibrosis. Using Elastica van Gieson-stained whole-slide images of pretreatment liver biopsies, collagen and elastin fibers were evaluated pixel by pixel (0.46 µm/pixel) using an automated computational method. Consequently, fiber amount and cumulative incidences of HCC within 3 years were analyzed. RESULTS: There was a significant correlation between collagen and elastin fibers, whereas variation in elastin fiber was greater than in collagen fiber. Both collagen fiber (p = 0.008) and elastin fiber (p < 0.001) were significantly correlated with F stage. In total, 30 patients developed HCC during follow-up. Patients who have higher elastin fiber (p = 0.002) in addition to higher collagen fiber (p = 0.05) showed significantly higher incidences of HCC. With regard to elastin fiber, this difference remained significant in F3 patients. Furthermore, for patients with a higher collagen fiber amount, higher elastin was a significant predictor for HCC development (p = 0.02). CONCLUSIONS: Computational analysis is a novel technique for quantification of fibers with the added value of conventional staging. Elastin fiber is a predictor for the development of HCC independently of collagen fiber and F stage.

Enhancing automatic classification of hepatocellular carcinoma images through image masking, tissue changes and trabecular features.

BACKGROUND: Recent breakthroughs in computer vision and digital microscopy have prompted the application of such technologies in cancer diagnosis, especially in histopathological image analysis. Earlier, an attempt to classify hepatocellular carcinoma images based on nuclear and structural features has been carried out on a set of surgical resected samples. Here, we proposed methods to enhance the process and improve the classification performance. METHODS: First, we segmented the histological components of the liver tissues and generated several masked images. By utilizing the masked images, some set of new features were introduced, producing three sets of features consisting nuclei, trabecular and tissue changes features. Furthermore, we extended the classification process by using biopsy resected samples in addition to the surgical samples. RESULTS: Experiments by using support vector machine (SVM) classifier with combinations of features and sample types showed that the proposed methods improve the classification rate in HCC detection for about 1-3%. Moreover, detection rate of low-grades cancer increased when the new features were appended in the classification process, although the rate was worsen in the case of undifferentiated tumors. CONCLUSIONS: The masking process increased the reliability of extracted nuclei features. The additional of new features improved the system especially for early HCC detection. Likewise, the combination of surgical and biopsy samples as training data could also improve the classification rates. Therefore, the methods will extend the support for pathologists in the HCC diagnosis.

Quantification of collagen and elastic fibers using whole-slide images of liver biopsy specimens.

Histological evaluation of fibrosis after a liver biopsy is crucial for evaluating the pathology of patients with chronic liver disease. Previous studies have reported quantitative analyses of fibrosis using images of collagen-stained sections. However, analysis of these studies requires manual selection of the region of interest. In addition, the quantification of elastic fibers is not considered. The present study was conducted in order to measure both the collagen and elastic fiber area ratios using Elastica van Gieson-stained whole-slide images (WSIs) of liver biopsy specimens. High-resolution WSIs provide precise color classification, enabling accurate detection of even fine collagen and elastic fibers. To minimize the influence of pre-existing fibrous tissue, median area ratios of the collagen and elastic fibers were independently calculated from the image tiles of the WSIs. These median area ratios were highly concordant with area ratios after the pre-existing fibrous tissues were manually trimmed from the WSI. Further, these median area ratios were correlated with liver stiffness as measured by transient elastography (collagen: r = 0.73 [P < 0.01], elastic: r = 0.53 [P < 0.01]). Our approach to quantifying liver fibrosis will serve as an effective tool to evaluate liver diseases in routine practice.

Computational grading of hepatocellular carcinoma using multifractal feature description.

Cancer grading has become an important topic in the field of image interpretation-based computer aided diagnosis systems. This paper proposes a novel feature descriptor to observe the characteristics of histopathological textures in a discriminative manner. The proposed feature descriptor utilizes fractal geometric analysis with four multifractal measures to construct an eight dimensional feature space. The proposed method employed a bag-of-feature-based classification model to discriminate a set of hepatocellular carcinoma images into five categories according to Edmondson and Steiner's grading system. Three feature selection methods were utilized to obtain the most discriminative features of codeword dictionary (codebook). Furthermore, we incorporated four other textural feature descriptors: Gabor-filters, LM-filters, local binary patterns, and Haralick, to obtain a benchmark of the accuracy of the classification. Two experiments were performed: (i) classifying non-neoplastic tissues and tumors and (ii) grading the hepatocellular carcinoma images into five classes. Experimental results indicated the significance of the multifractal features for describing the histopathological image texture because it outperformed other four feature descriptors. We graded a given ROI image by defining a threshold-based majority-voting rule and obtained an average correct classification rate around 95% for five classes classification.

Color correction for automatic fibrosis quantification in liver biopsy specimens.

CONTEXT: For a precise and objective quantification of liver fibrosis, quantitative evaluations through image analysis have been utilized. However, manual operations are required in most cases for extracting fiber areas because of color variation included in digital pathology images. AIMS: The purpose of this research is to propose a color correction method for whole slide images (WSIs) of Elastica van Gieson (EVG) stained liver biopsy tissue specimens and to realize automated operation of image analysis for fibrosis quantification. MATERIALS AND METHODS: Our experimental dataset consisted of 38 WSIs of liver biopsy specimens collected from 38 chronic viral hepatitis patients from multiple medical facilities, stained with EVG and scanned at ×20 using a Nano Zoomer 2.0 HT (Hamamatsu Photonics K.K., Hamamatsu, Japan). Color correction was performed by modifying the color distribution of a target WSI so as to fit to the reference, where the color distribution was modeled by a set of two triangle pyramids. Using color corrected WSIs; fibrosis quantification was performed based on tissue classification analysis. STATISTICAL ANALYSIS USED: Spearman's rank correlation coefficients were calculated between liver stiffness measured by transient elastography and median area ratio of collagen fibers calculated based on tissue classification results. RESULTS: Statistical analysis results showed a significant correlation r = 0.61-0.68 even when tissue classifiers were trained by using a subset of WSIs, while the correlation coefficients were reduced to r = 0.40-0.50 without color correction. CONCLUSIONS: Fibrosis quantification accompanied with the proposed color correction method could provide an objective evaluation tool for liver fibrosis, which complements semi-quantitative histologic evaluation systems.

Extraction of acne lesion in acne patients from multispectral images.

In acne treatment, it is important to accurately evaluate the severity of Acne. The acne should be classified into several skin lesions including comedo, reddish papule, pustule, and scar. However, in some cases, a visual detection from RGB image maybe difficult for the proper evaluation of acne skin lesions. This paper proposes an extraction method using the spectral information of the various type of acne skin lesions calculated from the multispectral images (MSI) of the lesions. In the experiment, we showed the possibility of classifying acne lesion types by applying a combination of several linear discriminant functions (LDF's).

Digital staining for multispectral images of pathological tissue specimens based on combined classification of spectral transmittance.

In this study, the digital transformation (digital staining) of the 16-band multispectral image of a hematoxylin and eosin (HE) stained pathological specimen to its Masson's trichrome (MT) stained counterpart is addressed. The digital staining procedure involves the classification of the various H&E-stained tissue components and then the transformation of their transmittance spectra to their equivalent MT-stained transmittance configurations. Combination of transmittance classifiers were designed to classify the various tissue components found in the multispectral images of an HE-stained specimen, e.g. nucleus, cytoplasm, red blood cell (RBC), fibrosis, etc.; while pseudo-inverse method was used to obtain the transformation matrices that would translate the transmittance spectra of the classified HE-stained multispectral pixels to their MT-stained configurations. To generate the digitally stained image, weighting factors, which were based on the classifiers beliefs, were introduced to the generated transformation matrices. Initial results of our experiments on liver specimens show the viability of multispectral imaging (MSI) to implement a digital staining framework in the pathological context.