This paper describes work presented at the Nordic Symposium on Digital Pathology 2015, in Linköping, Sweden. Prostatic intraepithelial neoplasia (PIN) represents premalignant tissue involving epithelial growth confined in the lumen of prostatic acini. In the attempts to understand oncogenesis in the human prostate, early neoplastic changes can be modeled in the mouse with genetic manipulation of certain tumor suppressor genes or oncogenes. As with many early pathological changes, the PIN lesions in the mouse prostate are macroscopically small, but microscopically spanning areas often larger than single high magnification focus fields in microscopy. This poses a challenge to utilize full potential of the data acquired in histological specimens. We use whole prostates fixed in molecular fixative PAXgene™, embedded in paraffin, sectioned through and stained with H&E. To visualize and analyze the microscopic information spanning whole mouse PIN (mPIN) lesions, we utilize automated whole slide scanning and stacked sections through the tissue. The region of interests is masked, and the masked areas are processed using a cascade of automated image analysis steps. The images are normalized in color space, after which exclusion of secretion areas and feature extraction is performed. Machine learning is utilized to build a model of early PIN lesions for determining the probability for histological changes based on the calculated features. We performed a feature-based analysis to mPIN lesions. First, a quantitative representation of over 100 features was built, including several features representing pathological changes in PIN, especially describing the spatial growth pattern of lesions in the prostate tissue. Furthermore, we built a classification model, which is able to align PIN lesions corresponding to grading by visual inspection to more advanced and mild lesions. The classifier allowed both determining the probability of early histological changes for uncategorized tissue samples and interpretation of the model parameters. Here, we develop quantitative image analysis pipeline to describe morphological changes in histological images. Even subtle changes in mPIN lesion characteristics can be described with feature analysis and machine learning. Constructing and using multidimensional feature data to represent histological changes enables richer analysis and interpretation of early pathological lesions.

Background: Numerous validation studies in digital pathology confirmed its value as a diagnostic tool. However, a longer time to diagnosis than traditional microscopy has been seen as a significant barrier to the routine use of digital pathology. As a part of our validation study, we compared a digital and microscopic diagnostic time in the routine diagnostic setting. Materials and Methods: One senior staff pathologist reported 400 consecutive cases in histology, nongynecological, and fine needle aspiration cytology (20 sessions, 20 cases/session), over 4 weeks. Complex, difficult, and rare cases were excluded from the study to reduce the bias. A primary diagnosis was digital, followed by traditional microscopy, 6 months later, with only request forms available for both. Microscopic slides were scanned at ×20, digital images accessed through the fully integrated laboratory information management system (LIMS) and viewed in the image viewer on double 23” displays. A median broadband speed was 299 Mbps. A diagnostic time was measured from the point slides were made available to the point diagnosis was made or additional investigations were deemed necessary, recorded independently in minutes/session and compared. Results: A digital diagnostic time was 1841 and microscopic 1956 min; digital being shorter than microscopic in 13 sessions. Four sessions with shorter microscopic diagnostic time included more cases requiring extensive use of magnifications over ×20. Diagnostic time was similar in three sessions. Conclusions: A diagnostic time in digital pathology can be shorter than traditional microscopy in the routine diagnostic setting, with adequate and stable network speeds, fully integrated LIMS and double displays as default parameters. This also related to better ergonomics, larger viewing field, and absence of physical slide handling, with effects on both diagnostic and nondiagnostic time. Differences with previous studies included a design, image size, number of cases, specimen type, network speed, and participant's level of confidence and experience in digital reporting. Further advancements in working stations and gained experience in digital reporting are expected to improve diagnostic time and widen routine applications of digital pathology.

Background: Intratumoral hypoxia plays an important role with regard to tumor biology and susceptibility to radio. and chemotherapy. For further investigation of hypoxia.related changes, areas of certain hypoxia must be reliably detected within cancer tissues. Pimonidazole, a 2.nitroimindazole, accumulates in hypoxic tissue and can be easily visualized using immunohistochemistry. Materials and Methods: To improve detection of highly hypoxic versus normoxic areas in prostate cancer, immunoreactivity of pimonidazole and a combination of known hypoxia.related proteins was used to create computational oxygen supply maps of prostate cancer. Pimonidazole was intravenously administered before radical prostatectomy in n = 15 patients, using the da Vinci robot.assisted surgical system. Prostatectomy specimens were immediately transferred into buffered formaldehyde, fixed overnight, and completely embedded in paraffin. Pimonidazole accumulation and hypoxia.related protein expression were visualized by immunohistochemistry. Oxygen supply maps were created using the normalized information from pimonidazole and hypoxia.related proteins. Results: Based on pimonidazole staining and other hypoxia.related proteins (osteopontin, hypoxia.inducible factor 1.alpha, and glucose transporter member 1) oxygen supply maps in prostate cancer were created. Overall, oxygen supply maps consisting of information from all hypoxia.related proteins showed high correlation and mutual information to the golden standard of pimonidazole. Here, we describe an improved computer.based ex vivo model for an accurate detection of oxygen supply in human prostate cancer tissue. Conclusions: This platform can be used for precise colocalization of novel candidate hypoxia.related proteins in a representative number of prostate cancer cases, and improve issues of single marker correlations. Furthermore, this study provides a source for further in situ tests and biochemical investigations

The Computer Science, Biology, and Biomedical Informatics (CoSBBI) program was initiated in 2011 to expose the critical role of informatics in biomedicine to talented high school students.^[1] By involving them in Science, Technology, Engineering, and Math (STEM) training at the high school level and providing mentorship and research opportunities throughout the formative years of their education, CoSBBI creates a research infrastructure designed to develop young informaticians. Our central premise is that the trajectory necessary to be an expert in the emerging fields of biomedical informatics and pathology informatics requires accelerated learning at an early age.In our 4^th year of CoSBBI as a part of the University of Pittsburgh Cancer Institute (UPCI) Academy (http://www.upci.upmc.edu/summeracademy/), and our 2nd year of CoSBBI as an independent informatics-based academy, we enhanced our classroom curriculum, added hands-on computer science instruction, and expanded research projects to include clinical informatics. We also conducted a qualitative evaluation of the program to identify areas that need improvement in order to achieve our goal of creating a pipeline of exceptionally well-trained applicants for both the disciplines of pathology informatics and biomedical informatics in the era of big data and personalized medicine.

Background: Intraductal proliferative lesions (IDPLs) of the breast are recognized as a risk factor for subsequent invasive carcinoma development. Although opportunities for IDPL diagnosis have increased, these lesions are difficult to diagnose correctly, especially atypical ductal hyperplasia (ADH) and low-grade ductal carcinoma in situ (LG-DCIS). In order to define the difference between these lesions, many molecular pathological approaches have been performed. However, still we do not have a molecular marker and objective histological index about IDPLs of the breast. Methods: We generated full digital pathology archives from 175 female IDPL patients, including usual ductal hyperplasia (UDH), ADH, LG-DCIS, intermediate-grade (IM)-DCIS, and high-grade (HG)-DCIS. After total 2,035,807 nucleic segmentations were extracted, we evaluated nuclear features using step-wise linear discriminant analysis (LDA) and a support vector machine. Results: High diagnostic accuracy (81.8–99.3%) was achieved between pathologists' diagnoses and two-group LDA predictions from nucleic features for IDPL discrimination. Grouping of nuclear features as size and shape-related or intranuclear texture-related revealed that the latter group was more important when distinguishing between normal duct, UDH, ADH, and LG-DCIS. However, these two groups were equally important when discriminating between LG-DCIS and HG-DCIS. The Mahalanobis distances between each group showed that the smallest distance values occurred between LG-DCIS and IM-DCIS and between ADH and Normal. On the other hand, the distance value between ADH and LG-DCIS was larger than this distance. Conclusions: In this study, we have presented a practical and useful digital pathological method that incorporates nuclear morphological and textural features for IDPL prediction. We expect that this novel algorithm is used for the automated diagnosis assisting system for breast cancer.