Accuracy in clinically evaluating pigmented lesions.

OBJECTIVE--To determine the ability of three doctors experienced in managing melanocytic lesions to diagnose correctly melanoma, dysplastic naevi, and various benign pigmented lesions. DESIGN--Independent clinical evaluation and histopathological assessment. SETTING--Pigmented lesion clinic, which patients attend without an appointment for early diagnosis of melanoma. PATIENTS--86 Patients with lesions that were judged to be benign by at least one of the three doctors. INTERVENTIONS--The lesions were excised under local anaesthesia and sent for histopathological examination in coded bottles without clinical details. MAIN OUTCOME MEASURE--Comparison of clinical with histopathological diagnosis for each lesion. RESULTS--A total of 120 lesions were evaluated by at least two of the three doctors. The histopathological diagnoses were made by the same pathologist. The overall sensitivity (diagnostic accuracy) for the three doctors for all types of lesion was 50%. Of the 39 dysplastic naevi, only 19 were identified correctly by all observers, and a further 24 banal lesions were wrongly diagnosed as dysplastic by at least one doctor. Particular difficulty was experienced with small (less than 5 mm), flat lesions, which can be banal or potentially malignant. CONCLUSIONS--Critical diagnosis and management decisions concerning pigmented lesions should always be based on a combination of clinical and histopathological assessments and the history of the patient.     

High resolution analysis of cervical cells--a progress report.

This paper presents preliminary results of research toward the development of a high resolution analysis stage for a dual resolution image processing-based prescreening device for cervical cytology. Experiments using both manual and automatic methods for cell segmentation are described. In both cases, 1500 cervical cells were analyzed and classified as normal or abnormal (dysplastic or malignant) using a minimum Mahalanobis distance classifier with eight subclasses of normal cells, and five subclasses of abnormal cells. With manual segmentation, false positive and false negative error rates of 2.98 and 7.73% were obtained. Similar experiments using automatic cell segmentation methods yielded false positive and false negative error rates of 3.90 and 11.56%, respectively. In both cases, independent training and testing data were used.     

Phagocytosis of autologous lymphocytes by cervical preneoplastic and neoplastic cells

Eighty-nine random Pap smears of the uterine cervix were examined to evaluate the phagocytic abnormal cells (PACs) of atypical, dysplastic and neoplastic tissues against infiltrated blood cells. The results revealed that none of the PACs have been identified in atypical (II) and mild dysplastic cells (IIIa). Low levels (1.2%) of PACs were initially demonstrated in patients with moderate dysplasia (IIIb) that increased 1.7-fold in subsequent severe dysplasia (IIIc) and further increased 2.8-fold in invasive carcinomas of the cervix (V). In addition, the data showed age-related responsiveness toward the development of PACs, where 82% of old patients have developed phagocytic activity in moderate dysplastic cells compared with 27% of young patients. However, the difference became less significant at the subsequent classes of the disease. These data further demonstrated that PACs are class-dependent and it may explain one mechanism by which precancerous cells escape immunosurveillance.     

Flow sorting of tumor cells for morphometric analysis, particularly of rare cells.

Using flow cytometric DNA measurement and sorting combined with morphometric light microscopy, different groups of cells were studied in a human melanoma pleural effusion, a human melanoma lymph node metastasis and a mouse tumor, as well as in normal reference tissues. Beside cells of the predominant tumor cell population, three types of rare tumor cells were studied after enrichment by sorting: a) giant cells from the greater than 8c region, comprising about 5% of the tumor cells, b) binucleated and multinucleated cells with unequal nuclear sizes within the same cell, found at frequencies of about 1.5%, and c) less than 2c cells which were derived from the so-called "debris"-region of the DNA histogram, found at frequencies of about 1 to 6%. All these rare cells were found only in the malignant tumors and not in the benign reference tissues. Morphometry showed that the increase in the cellular DNA content in the different fractions of tumor cells was combined with an increase in the cellular and nuclear sizes. However, the n/c-ratio was constant in the whole range of tumor cell fractions, including the fractions from the the less than 2c and the greater than 8c regions. The n/c-ratio of the less than 2c cells and giant cells differed from that of corresponding normal cells underlining their origin from the predominant tumor cell population. The possible linkage between the occurrence of the three rare cell types and genetic instability of tumors related to faulty nucleus and cell division is discussed.     

A rapid screening classifier for diagnosing COVID-19

Rationale: Coronavirus disease 2019 (COVID-19) has caused a global pandemic. A classifier combining chest X-ray (CXR) with clinical features may serve as a rapid screening approach.Methods: The study included 512 patients with COVID-19 and 106 with influenza A/B pneumonia. A deep neural network (DNN) was applied, and deep features derived from CXR and clinical findings formed fused features for diagnosis prediction.Results: The clinical features of COVID-19 and influenza showed different patterns. Patients with COVID-19 experienced less fever, more diarrhea, and more salient hypercoagulability. Classifiers constructed using the clinical features or CXR had an area under the receiver operating curve (AUC) of 0.909 and 0.919, respectively. The diagnostic efficacy of the classifier combining the clinical features and CXR was dramatically improved and the AUC was 0.952 with 91.5% sensitivity and 81.2% specificity. Moreover, combined classifier was functional in both severe and non-serve COVID-19, with an AUC of 0.971 with 96.9% sensitivity in non-severe cases, which was on par with the computed tomography (CT)-based classifier, but had relatively inferior efficacy in severe cases compared to CT. In extension, we performed a reader study involving three experienced pulmonary physicians, artificial intelligence (AI) system demonstrated superiority in turn-around time and diagnostic accuracy compared with experienced pulmonary physicians.Conclusions: The classifier constructed using clinical and CXR features is efficient, economical, and radiation safe for distinguishing COVID-19 from influenza A/B pneumonia, serving as an ideal rapid screening tool during the COVID-19 pandemic.     

Rab11 in dysplasia of Barrett's epithelia

Barrett's esophagus predisposes affected patients to the development of esophageal adenocarcinoma. The development of adenocarcinoma proceeds along a progression through low- and high-grade dysplasia. Surveillance of Barrett's patients requires serial endoscopic investigations and grading mucosal biopsies. Unfortunately, grading of biopsies by conventional hematoxylin and eosin staining is fraught with significant interobserver variations. We have found in both biopsy and resection specimens that immunostaining for the small GTP binding protein Rab11 is increased in low-grade dysplastic cells. This staining is lost in high-grade dysplastic cells. These results suggest that low-grade dysplastic cells undergo an apical trafficking blockade, which is released as cells progress to the less differentiated phenotype of high-grade dysplasia and adenocarcinoma. Examination of the SKGT-4 esophageal adenocarcinoma cell line demonstrated prominent mRNA and protein expression for Rab11. Rab11 immunostaining was present in SKGT-4 cells as a perinuclear nidus of punctate staining along with a more diffuse punctate pattern. Thus, Rab11 expression was present in a esophageal adenocarcinoma cells in culture. Markers of vesicle trafficking may be critical factors for grading of mucosal dysplastic transitions leading to adenocarcinoma.     

Dysplastic nevus--risk factor or disguise for melanoma

Dysplastic nevus is an acquired or hereditary nevus that clinically seems atypical and pathohistologically dysplastic. The term of dysplastic nevus has changed through history and even until now the dermatologists and pathologists have not found the same conclusion for name and definition of dysplastic nevus. Epidemiology of dysplastic nevus is different depending on geographic lattitude, being three times higher in Australia than in Great Britain. Genetic factors play a role in etiology of dysplastic nevus but are still not well defined. UV radiation is indisputable main etiological factor in developing dysplastic nevus. Many studies confirm that children who have been using sun protection creams with SPF have less dysplastic nevi than those who did not. Nevus with geographic shape and muddy borders, dominately macular, red to brown colored and has 5 mm or more in diameter is clinically dysplastic nevus. ABCDE rules count for dysplastic nevus as well as for melanoma but prefferable diagnostic criteria for dysplastic nevus would be "ugly duckling sign". Pathohistologic analysis is the key in confirming the diagnosis of dysplastic nevus. Great experience and knowledge in dermatopathology field is essential for pathologists to make a distinction between dysplastic nevus and melanoma in situ. Likewise great experience in dermatooncology field is essential in differentiating dysplastic nevus from other nevi. Surgical excision is the only therapy that should be done for dysplastic nevus. Regular follow up is highly recommended for patients with dysplastic nevus and syndroma naevi dysplastic. Education about sun protection measures and self-examination techniques is essential for all patients with dysplastic nevi and their family.     

The Anal Pap Smear: Cytomorphology of squamous intraepithelial lesions

BACKGROUND: Anal smears are increasingly being used as a screening test for anal squamous intraepithelial lesions (ASILs). This study was undertaken to assess the usefulness and limitations of anal smears in screening for ASILs. METHODS: The cytomorphological features of 200 consecutive anal smears collected in liquid medium from 198 patients were studied and findings were correlated with results of surgical biopsies and/or repeat smears that became available for 71 patients within six months. RESULTS: Adequate cellularity was defined as an average of 6 or more nucleated squamous cells/hpf. A glandular/transitional component was not required for adequacy. Dysplastic cells, atypical parakeratotic cells and bi/multinucleated cells were frequent findings in ASIL while koilocytes were infrequent. Smears from LSIL cases most frequently showed mildly dysplastic and bi/multinucleate squamous cells followed by parakeratotic cells (PK), atypical parakeratotic cells (APK), and koilocytes. HSIL smears contained squamous cells with features of moderate/severe dysplasia and many APKs. Features of LSIL were also found in most HSIL smears. CONCLUSIONS: In this study liquid based anal smears had a high sensitivity (98%) for detection of ASIL but a low specificity (50%) for predicting the severity of the abnormality in subsequent biopsy. Patients with cytologic diagnoses of ASC-US and LSIL had a significant risk (46-56%) of HSIL at biopsy. We suggest that all patients with a diagnosis of ASC-US and above be recommended for high resolution anoscopy with biopsy.     

Single-cell DNA measurement in hyperplastic, dysplastic and carcinomatous laryngeal epithelia, with special reference to the occurrence of hypertetraploid cell nuclei

Single-cell DNA fluorometry was performed on smear preparations from 57 histologically verified lesions from the vocal cords. Special attention was paid to the detection of hypertetraploid cell nuclei, which may be considered neoplastic markers, since they are not present in normal vocal cord epithelium. Hypertetraploid nuclei were not found in hyperplastic or mildly dysplastic epithelia but were found in 4 of 14 specimens of moderate dysplasia and in 5 of 19 specimens of severe dysplasia. The presence of hypertetraploid nuclei was associated with more frequent recurrences and progression of disease. Hypertetraploid cell nuclei were found in 9 of 15 invasive carcinomas and in this group were associated with a worse prognosis. The proliferative activity was 1.7% for hyperplasia and mild dysplasia, 3.1% and 2.8%, respectively, for moderate and severe dysplasia and 7.6% for squamous-cell carcinoma.     

Effects of an inhibitor of tripeptidyl peptidase II (Ala-Ala-Phe-chloromethylketone) and its combination with an inhibitor of the chymotrypsin-like activity of the proteasome (PSI) on apoptosis, cell cycle and proteasome activity in U937 cells

AAF-AMC is not a specific TPP II substrate, since it is also hydrolyzed by purified proteasomes. Moreover, AAF-cmk, claimed to be a specific TPP II inhibitor, also inhibits the chymotrypsin-like activity of the proteasome. While AAF-cmk itself is mildly cytostatic to U-937 cells and induces cell cycle block in G1, its combination with PSI does not induce an increase in the cytostatic/cytotoxic effects. This suggests that TPP II is possibly less important for cell metabolism than it was previously believed and it is less probable that it can be able to fully compensate for the loss of the proteasome function.     

Overlapped Partitioning for Ensemble Classifiers of P300-Based Brain-Computer Interfaces

A P300-based brain-computer interface (BCI) enables a wide range of people to control devices that improve their quality of life. Ensemble classifiers with naive partitioning were recently applied to the P300-based BCI and these classification performances were assessed. However, they were usually trained on a large amount of training data (e.g., 15300). In this study, we evaluated ensemble linear discriminant analysis (LDA) classifiers with a newly proposed overlapped partitioning method using 900 training data. In addition, the classification performances of the ensemble classifier with naive partitioning and a single LDA classifier were compared. One of three conditions for dimension reduction was applied: the stepwise method, principal component analysis (PCA), or none. The results show that an ensemble stepwise LDA (SWLDA) classifier with overlapped partitioning achieved a better performance than the commonly used single SWLDA classifier and an ensemble SWLDA classifier with naive partitioning. This result implies that the performance of the SWLDA is improved by overlapped partitioning and the ensemble classifier with overlapped partitioning requires less training data than that with naive partitioning. This study contributes towards reducing the required amount of training data and achieving better classification performance.     

Atelencephalic microcephaly in a 21-week human fetus

Atelencephalic microcephaly, a rare and extreme disorder, is known morphologically by only six cases. Derivatives of the telencephalon are absent or dysplastic, while more caudal structures are normal or mildly deformed. A more extensive form, aprosencephaly, involves structures of the diencephalon and may be associated with holoprosencephalic facies. Extracranial anomalies may be present in both atelencephaly and aprosencephaly. We describe the seventh and youngest specimen, a 21-week female with atelencephaly. Maternal and gestational histories were unremarkable; the fetus was obtained by therapeutic abortion following diagnosis of a severe cranial malformation by ultrasound. A small and depressed, but intact, calvarium covered the brain. The forebrain was rounded and showed fused hemispheres, absent gyri, olfactory bulbs, and tracts. Caudal structures were mildly deformed. An oval mass of bone filled much of the middle cranial fossa. By light microscopy, several dysplastic changes were apparent in the forebrain. Ventricles were not present; small round cells resembling those of the germinal matrix were prominent in the forebrain. Pyramidal tracts were absent at all levels. The process responsible for these changes cannot be established with certainty; the changes are, however, in keeping with previous damage, such as that accepted for other encephaloclastic disorders. The insult in atelencephaly presumably occurs after closure of the rostral neuropore; earlier damage, with more widespread consequences, is possible for aprosencephaly. As with other destructive processes, etiology in atelencephaly and aprosencephaly is most likely heterogeneous.     

Identification and optimization of classifier genes from multi-class earthworm microarray dataset

Monitoring, assessment and prediction of environmental risks that chemicals pose demand rapid and accurate diagnostic assays. A variety of toxicological effects have been associated with explosive compounds TNT and RDX. One important goal of microarray experiments is to discover novel biomarkers for toxicity evaluation. We have developed an earthworm microarray containing 15,208 unique oligo probes and have used it to profile gene expression in 248 earthworms exposed to TNT, RDX or neither. We assembled a new machine learning pipeline consisting of several well-established feature filtering/selection and classification techniques to analyze the 248-array dataset in order to construct classifier models that can separate earthworm samples into three groups: control, TNT-treated, and RDX-treated. First, a total of 869 genes differentially expressed in response to TNT or RDX exposure were identified using a univariate statistical algorithm of class comparison. Then, decision tree-based algorithms were applied to select a subset of 354 classifier genes, which were ranked by their overall weight of significance. A multiclass support vector machine (MC-SVM) method and an unsupervised K-mean clustering method were applied to independently refine the classifier, producing a smaller subset of 39 and 30 classifier genes, separately, with 11 common genes being potential biomarkers. The combined 58 genes were considered the refined subset and used to build MC-SVM and clustering models with classification accuracy of 83.5% and 56.9%, respectively. This study demonstrates that the machine learning approach can be used to identify and optimize a small subset of classifier/biomarker genes from high dimensional datasets and generate classification models of acceptable precision for multiple classes.     

How often is dysplasia diagnosed by biopsy or smear examination? Application of a maximum likelihood based method to the assessment of detection rates in the nasal mucosa of nickel workers.

In view of the known increased risk of nasal carcinoma and the high prevalence of dysplastic lesions of the nasal mucosa among nickel workers, regular screening for the existence of possibly precancerous dysplastic lesions is offered to workers in a Norwegian nickel refinery. Unfortunately, available sampling techniques do not allow the identification of all subjects in whom dysplastic changes are present. Independent histological and cytological (brush cytology) diagnoses, obtained for each of a group of 90 workers, have been used to estimate, by a maximum likelihood method, the probabilities that existing dysplastic lesions will be detected by each of these two screening methods. In the group studied, cytology performed rather less well than histology in unambiguously detecting dysplasia. However, when cytological specimens showing irregular (possibly dysplastic) epithelial cells were grouped with those showing clear dysplastic changes, detection probabilities were estimated at 0.52 by histology and 0.57 by cytology. Detection probabilities were estimated to be higher among subjects with a previously known history of dysplasia, particularly by histology (P < 0.01), probably due to larger dysplastic areas. In view of both its greater facility and speed of sampling, and its greater acceptability, brush cytology may be preferable to biopsy sampling for the screening of large numbers of workers at risk.     

Studies with the Golgi method in central gangliogliomas and dysplastic gangliocytoma of the cerebellum (Lhermitte-Duclos disease).

The rapid Golgi method, combined with current optical and electronmicroscopical techniques, was used in three central gangliogliomas and in one dysplastic gangliocytoma of the cerebellum to study the morphology of ganglionic cells. Gangliogliomas were composed of bipolar, fusiform and radiate cells with dense core and clear vesicles in the perikaryon and cellular processes, the number of each cellular type varying from one case to another. These features, together with the fact that isodendritic neurons are considered to be phylogenetically old neurons, suggest that these tumours are composed of "primitive" neurons that are not homogeneous with regard to their morphology. In contrast, ganglionic cells in dysplastic gangliocytoma are huge cells with long, stereotyped neurites that establish unique asymmetric contacts with neighbouring perikarya and neurites by means of claw-shaped processes covered with synaptic buttons. These morphological characteristics are different from those of any other neuron of the CNS.     

ProLoc: prediction of protein subnuclear localization using SVM with automatic selection from physicochemical composition features

Accurate prediction methods of protein subnuclear localizations rely on the cooperation between informative features and classifier design. Support vector machine (SVM) based learning methods are shown effective for predictions of protein subcellular and subnuclear localizations. This study proposes an evolutionary support vector machine (ESVM) based classifier with automatic selection from a large set of physicochemical composition (PCC) features to design an accurate system for predicting protein subnuclear localization, named ProLoc. ESVM using an inheritable genetic algorithm combined with SVM can automatically determine the best number m of PCC features and identify m out of 526 PCC features simultaneously. To evaluate ESVM, this study uses two datasets SNL6 and SNL9, which have 504 proteins localized in 6 subnuclear compartments and 370 proteins localized in 9 subnuclear compartments. Using a leave-one-out cross-validation, ProLoc utilizing the selected m=33 and 28 PCC features has accuracies of 56.37% for SNL6 and 72.82% for SNL9, which are better than 51.4% for the SVM-based system using k-peptide composition features applied on SNL6, and 64.32% for an optimized evidence-theoretic k-nearest neighbor classifier utilizing pseudo amino acid composition applied on SNL9, respectively.     

Computerized morphometric discrimination between normal and tumoral cells in oral smears

The oral exfoliative cytology allows a quick and fairly accurate assessment of suspicious lesions of the oral cavity. Within this context, our paper proposes a quantitative approach, focusing on the construction of a classifier for detecting the presence of the tumoral cells on oral smears. The design of the classifier relies on a detailed computerized analysis of the individual morphometric features exhibited by two large known populations of normal and tumoral cells, respectively; the digital image processing was performed in the Zeiss KS400 environment. The classifier was implemented as a neural network with step activation function, whose parameters were obtained from an adequate training, based on the nuclear and cytoplasmic areas of the cells belonging to the two populations. Our procedure based on this classifier was meant to operate by identifying the tumoral or normal nature of any cell randomly selected from a smear. To identify the nature of an arbitrary cell, its nuclear and cytoplasmic areas are presented at the input of the classifier. The classification procedure was tested on several smears, and the results coincided with the pathological diagnosis in all the considered cases. The performances of our approach are discussed in comparison with other analytical methods previously reported in oral exfoliative cytology. These discussions emphasize the role of numerical information exploited for the classifier design, concluding that the individual morphometric features are more meaningful than the global characterization of smears by mean values.     

Using Chou’s pseudo amino acid composition based on approximate entropy and an ensemble of AdaBoost classifiers to predict protein subnuclear location

The knowledge of subnuclear localization in eukaryotic cells is essential for understanding the life function of nucleus. Developing prediction methods and tools for proteins subnuclear localization become important research fields in protein science for special characteristics in cell nuclear. In this study, a novel approach has been proposed to predict protein subnuclear localization. Sample of protein is represented by Pseudo Amino Acid (PseAA) composition based on approximate entropy (ApEn) concept, which reflects the complexity of time series. A novel ensemble classifier is designed incorporating three AdaBoost classifiers. The base classifier algorithms in three AdaBoost are decision stumps, fuzzy K nearest neighbors classifier, and radial basis-support vector machines, respectively. Different PseAA compositions are used as input data of different AdaBoost classifier in ensemble. Genetic algorithm is used to optimize the dimension and weight factor of PseAA composition. Two datasets often used in published works are used to validate the performance of the proposed approach. The obtained results of Jackknife cross-validation test are higher and more balance than them of other methods on same datasets. The promising results indicate that the proposed approach is effective and practical. It might become a useful tool in protein subnuclear localization. The software in Matlab and supplementary materials are available freely by contacting the corresponding author.