Machine Learning in the Detection of Oral Lesions With Clinical Intraoral Images.

INTRODUCTION: Artificial intelligence in oncology has gained a lot of interest in recent years. Early detection of Oral squamous cell carcinoma (OSCC) is crucial for early management to attain a better prognosis and overall survival. Machine learning (ML) has also been used in oral cancer studies to explore the discrimination between clinically normal and oral cancer. MATERIALS AND METHODS: A dataset comprising 360 clinical intra-oral images of OSCC, Oral Potentially Malignant Disorders (OPMDs) and clinically healthy oral mucosa were used. Clinicians trained the machine learning model with the clinical images (n=300). Roboflow software (Roboflow Inc, USA) was used to classify and annotate images along with Multi-class annotation and object detection models trained by two expert oral pathologists. The test dataset (n=60) of new clinical images was again evaluated by two clinicians and Roboflow. The results were tabulated and Kappa statistics was performed using SPSS v23.0 (IBM Corp., Armonk, NY).  Results: Training dataset clinical images (n=300) were used to train the clinicians and Roboflow algorithm. The test dataset (n=60) of new clinical images was again evaluated by the clinicians and Roboflow. The observed outcomes revealed that the Mean Average Precision (mAP) was 25.4%, precision 29.8% and Recall 32.9%. Based on the kappa statistical analysis the 0.7 value shows a moderate agreement between the clinicians and the machine learning model. The test dataset showed the specificity and sensitivity of the Roboflow machine learning model to be 75% and 88.9% respectively.  Conclusion: In conclusion, machine learning showed promising results in the early detection of suspected lesions using clinical intraoral images and aids general dentists and patients in the detection of suspected lesions such as OPMDs and OSCC that require biopsy and immediate treatment.

It's the data, stupid: Inflection point for Artificial Intelligence in Indian healthcare.

Indian healthcare is fast growing and with significant chunk of it being in small, fragmented, informal sector; Artificial Intelligence (AI) is pegged as a magical tool for a better healthcare system. There is an inclination to merely mimic the US approach in the on-going policy making and legislative exercises, which can have serious fallouts for Indian healthcare. India needs a different approach to suite her unique requirements. In this regard, each of the five stages in AI development lifecycle has been analyzed in the light of current on-ground realities. These boil down to three fold challenges of how to increase adoption of digital health, prevent data silos and create maximum value from data. Availability of quality data for value addition without barriers and restrictions is the common denominator for leveraging the full potential of AI. This requires liberal policies enabling secondary use of data in developing countries with rapidly growing healthcare sector akin to India. This has to be carefully balanced with data privacy and security. Restrictive healthcare data policies and laws can slow down adoption of digitization, perpetuate status-quo, be biased towards the incumbent players, cause Industry stagnation and thus will do more harm than good. It is therefore the data policies that will make or break AI in Indian healthcare.

Prosthetic management of posttraumatic external auditory canal atresia: A rare cause of conductive hearing loss.

Acquired external auditory canal (EAC) atresia is an infrequent entity which can originate from a number of different causes including trauma, infection, neoplasia, inflammation, and radiotherapy. Posttraumatic atresias are exceptionally rare, only 10% of atresias are attributed to trauma in most of the series. The management of stenosis of the EAC is challenging as it is associated with residual hearing loss and late recurrence. Traditional stents often occlude the EAC, resulting in a temporary conductive hearing loss. This case report describes the technique of fabrication of a wide-bored acrylic stent which attained additional retention from the folds of the auricle. The customized earmold stent effectively prevented restenosis, while the large bore provided ventilation and improved hearing subjectively during the stenting period.

Phyllodes tumor: a clinicopathologic and immunohistochemical study of 30 cases.

CONTEXT: Phyllodes tumors (PTs) of the breast are biphasic neoplasms composed of epithelium and a spindle-cell stroma. Currently, PTs are classified as benign, borderline, or malignant based on histopathologic features. However, histologic classification does not always predict outcome. Objective.-To determine the prognostic value of a variety of clinicopathologic features and immunoreactivities in PTs. DESIGN: Sixteen benign, 8 borderline, and 6 malignant PTs with follow-up were examined for reactivity across a panel of immunohistochemical stains, including c-Kit, endothelin 1, p16, p21, p53, and Ki-67. Clinicopathologic features, including stromal cellularity, mitotic rate, and margin status, were also assessed. Tumor variables were compared among tumor subgroups and between tumors that did and did not recur. RESULTS: Of the 30 PTs, 4 recurred (1 benign, 2 borderline, 1 malignant). One patient with a malignant tumor died of metastatic disease 34 months after initial diagnosis. The overall positive rate of c-Kit immunoreactivity was 13% in benign, 63% in borderline, and 67% in malignant PTs. Endothelin 1 epithelial cytoplasmic staining was seen in 100% of benign, 50% of borderline, and 17% of malignant PTs. Additionally, p16, p21, p53, and Ki-67 were differentially expressed among benign, borderline, and malignant tumors. Positive surgical resection margins was the only variable that significantly predicted recurrent disease (P = .02). CONCLUSIONS: Stromal c-Kit positivity and epithelial endothelin 1 negativity are more often associated with malignant PTs; however, only positive margin status is significantly associated with tumor behavior.

Serum phosphate and lactate vary with FSH in an early postmenopausal population.

OBJECTIVE: We examined serum in recent postmenopausal women to determine the relationship of menopausal status, as FSH level, to serum acid-base balance. DESIGN, SETTING, AND PATIENTS: Serum electrolytes of 58 women, aged 53-58, were measured relative to serum FSH. The subjects were over one year since the last menstrual period and were from an academic practice setting. RESULTS: In women with FSH <35 IU/L (n=20, mean 16.6 IU/L, SD 6.8), phosphate and lactate were reduced relative to women with FSH >35 IU/L (n=38, mean 84.8 IU/L, SD 34.5). No other major anions showed significant differences. Both groups were analyzed by mass spectroscopy for fatty acids and anionic metabolic intermediates. Lactate was the predominant anion in the organic group but accounted for only about 10% of the FSH-responsive anion change. This change was mainly due to a 0.1-mM increase in phosphate in the high FSH group. CONCLUSIONS: There is a direct correlation of early postmenopausal FSH to increasing serum phosphate. Changes in phosphorus may reflect differences in the rate of bone loss.

Microdissection genotyping of mixed glial and primitive neuroectodermal central nervous system neoplasm.

A 22-year-old man with previous radiation treatment for childhood astrocytoma underwent resection of a right parietooccipital lesion. Histopathology revealed a malignant neoplasm with areas of astrocytic and primitive neuroectodermal components. To resolve the relationship and cellular origin, representative tissue was microdissected from several targets, obtaining a balanced mixture of each element. Nonneoplastic brain parenchyma was separately microdissected to determine polymorphic marker informativeness and to serve as an internal negative control. Despite the relatively small quantity of tissue removed for each microdissection target, sufficient material was available for reliable, balanced, polymerase chain reaction-format genotyping encompassing a panel of tumor suppressor genes and genetic loci associated with these forms of neoplasia. The findings revealed distinct discordant genotypic profiles for each of the neoplastic components. The efficacy of the approach used for molecular analysis of this complex neoplasm and the implication of the genotypic findings are discussed.

A case of breathlessness, confusion and diarrhoea.

A 61-year-old woman presented to the emergency department with acute-onset breathlessness, fever, sore throat and confusion. Her initial investigations revealed hyponatremia and middle lobe consolidation. In view of the atypical symptoms and signs, erythromycin was commenced. Urinary legionella antigen was requested and that tested positive. She was one of the first few patients whose findings alerted us to a possible outbreak of legionnaire' disease. We drew the following conclusions from our experience with this and other cases that we saw during the legionnaires' outbreak: an atypical presentation is common, as seen in this lady with confusion. If two cases of atypical pneumonias test positive for legionella within a period of a week, we suggest that public health authorities should be notified to identify the source and contain it. There is a need for continuous and high vigilance for legionnaires' disease.

Microdissection genotyping of gliomas: therapeutic and prognostic considerations.

Molecular anatomic pathology represents the blend of traditional morphological methods and the multigene approach to determine cancer-related gene alterations for diagnostic and prognostic purposes. Microdissection genotyping was utilized to characterize 197 gliomas with targeted microdissection of 2-7 areas spanning the spectrum of histologic types and grades. The methodology described herein is complementary to the existing realities of pathology practice. The technique utilizes paraffin-embedded fixative-treated tissue of small sample size after the primary morphological examination by the pathologist. Molecular information derived from microdissection genotyping in combination with the traditional histological information, results in an enhanced understanding of glioma formation and biological progression leading to improvements in diagnosis and prediction of prognosis. In all, 100% or 32 of 32 cases with at least partial treatment response was observed in neoplasms possessing the 1p or 1p/19q loss. The 19q loss alone without coexisting 1p showed no improvement in treatment response. Gliomas lacking 1p loss with only allelic loss involving 3p, 5q, 9p, 10q and 17p showed unfavorable outcome of only 35%, or six of 17 cases with treatment response. In addition, the determination of fractional allelic loss (favorable/unfavorable), was a very good independent predictor of biological behavior. These findings emphasize the importance of determining the cumulative pattern of mutational damage on 16 distinct sites or more, especially in the presence of 1p loss which in isolation or in combination with 19q is a favorable prognostic factor for therapeutic response.

Microdissection-based genotyping assists discrimination of reactive gliosis from glioma.

We used molecular anatomic pathology to determine the mutational status (loss of heterozygosity [LOH]) to make the distinction between reactive gliosis and glial neoplasia. LOH has been shown to be absent in reactive states and present in neoplastic cellular proliferations. Three groups of patient specimens were analyzed: group 1, reactive gliosis (n = 15); group 2, gliomas of varying histologic type and grade (n = 54); group 3, diagnostically challenging reactive gliosis vs glioma (n = 16). No group 1 cases (0/15 [0%]) showed allelic loss, whereas all high-grade glial neoplasms, a subset of group 2 (35/35 [100%]) manifested at least 1 allelic loss alteration, with most cases (33/35 [94%]) displaying 2 or more such changes. During a look forward at the group 3 patients, clinical history clarified the problematic diagnosis in a subset of 11 patients: 8 (73%) of 11 clinical outcomes were predicted correctly by our analysis. The molecular anatomic pathology approach outlined herein is designed for minute, formalin-fixed, paraffin-embedded specimens, which are encountered in everyday surgical pathology practice. Molecular anatomic pathology opens the possibilities of molecular analysis to everyday pathology practice.

Microdissection genotyping of archival fixative treated tissue for Gaucher disease.

The genetic diagnosis of Gaucher disease by molecular methods is complicated by the existence of a highly homologous transcribed pseudogene (96% identity) that is found in close proximity to the true gene on chromosome 1q21. In addition, the pseudogene sequence can mimic disease-causing mutations in the true gene. Selective polymerase chain reaction (PCR) amplification of the true gene can be accomplished in extracted DNA from fresh-frozen samples by designing oligonucleotide primers to hybridize to defined regions that are not present in the pseudogene. This standard molecular approach, which entails amplification of relatively long segments of intact DNA, is not feasible in archival, paraffin-embedded, solid-tissue specimens in which the negative effects of chemical fixation result in DNA strand scission and breakdown of nucleic acid. A novel approach, specifically created for use with archival, fixative-treated tissue specimens, was developed for detection and characterization of common mutations of Gaucher disease. Three separate robust PCR reactions were formulated, 2 for selective amplification of portions of only the true gene exons 2 and 9, with a third reaction targeting exon 10, wherein both the true and pseudogene were coamplified. In the latter, DNA sequencing was used to determine the presence of true and pseudogene allele content in addition to identification of base sequence alterations. This method, requiring a single, 4-microm-thick histologic section, was successfully applied to archival paraffin block tissue specimens that had been in storage for up to 75 years. It was capable of accurately genotyping common Gaucher disease mutations as well as discovering a novel mutation and genetic polymorphism. We recommend our approach when only fixative-treated tis sue is available for molecular genotyping.