Artificial intelligence for classification and detection of oral mucosa lesions on photographs: a systematic review and meta-analysis.

OBJECTIVE: This study aimed to review and synthesize studies using artificial intelligence (AI) for classifying, detecting, or segmenting oral mucosal lesions on photographs. MATERIALS AND METHOD: Inclusion criteria were (1) studies employing AI to (2) classify, detect, or segment oral mucosa lesions, (3) on oral photographs of human subjects. Included studies were assessed for risk of bias using Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2). A PubMed, Scopus, Embase, Web of Science, IEEE, arXiv, medRxiv, and grey literature (Google Scholar) search was conducted until June 2023, without language limitation. RESULTS: After initial searching, 36 eligible studies (from 8734 identified records) were included. Based on QUADAS-2, only 7% of studies were at low risk of bias for all domains. Studies employed different AI models and reported a wide range of outcomes and metrics. The accuracy of AI for detecting oral mucosal lesions ranged from 74 to 100%, while that for clinicians un-aided by AI ranged from 61 to 98%. Pooled diagnostic odds ratio for studies which evaluated AI for diagnosing or discriminating potentially malignant lesions was 155 (95% confidence interval 23-1019), while that for cancerous lesions was 114 (59-221). CONCLUSIONS: AI may assist in oral mucosa lesion screening while the expected accuracy gains or further health benefits remain unclear so far. CLINICAL RELEVANCE: Artificial intelligence assists oral mucosa lesion screening and may foster more targeted testing and referral in the hands of non-specialist providers, for example. So far, it remains unclear if accuracy gains compared with specialized can be realized.

Attentional network changes in subjective cognitive decline.

BACKGROUND: Subjective cognitive decline (SCD) is known as the intermediate stage between normal cognitive aging and mild cognitive impairment (MCI). Although elderly with SCD usually perform close to normal in standardized tests, the detailed function of attention networks in this group has not been studied yet. AIMS: The purpose of this study was to investigate the performance of attention networks, as a possible indicator of cognitive disorder, in older individuals with subjective memory complaint and MCI. METHOD: The attention network test (ANT) was used to examine and compare the performance of three attention networks of alerting, orientation, and executive control in 17 elderly with SCD, 30 multiple domain amnestic MCI subjects, and 15 healthy controls. RESULTS: Although the orienting network had almost the same performance in all groups (p = 0.25), the performance of alerting (p = 0.01) and executive control networks (p = 0.02) were significantly different among the three groups: the SCD group performed poorly in both networks compared with the controls and did not differ significantly from the MCI group (p ≥ 0.05). However, controlling for general age-related slowing abolished the group difference in executive control index. More importantly, our results showed that alerting network that was affected in SCD group had high sensitivity in differentiating this group from controls (0.94%). CONCLUSION: Our data suggest that despite normal performance in neuropsychological tests, the SCD elderly may face significant degrees of attention processing problems, especially in maintaining alerting to external stimuli which might be helpful in diagnosing individuals at risk and designing proper attention-based interventions.

Executive function deficits in mild cognitive impairment: evidence from saccade tasks.

OBJECTIVES: Early detection of mild cognitive impairment (MCI) is necessary to prevent irreversible brain damage caused by incipient Alzheimer's disease. It has been showing that amnestic MCI (a-MCI) subjects exhibit subtle deficits in executive function that can be tested using saccade eye movements. Eye-tracking technology is a sensitive method to measure cognitive impairments in dementia and MCI. METHODS: In this study, we used eye-tracking technology to explore saccade impairments to distinguish between a-MCI and the variants of reference controls. 21 patients with AD, 40 patients with a-MCI, and 59 normal participants were recruited in current study. We measured saccade reaction time, saccade errors, saccade omission, and uncorrected saccades using anti-saccade and pro-saccade tasks with 'gap' and 'overlap' procedures. These parameters were used as markers of executive function and visual attention deficits.Results: The findings revealed that more errors, more omissions, and fewer corrections characterized the saccade behavior of the a-MCI group compared to the reference group. These eye-tracking characteristics can be considered as inhibitory control and working memory deficits in a-MCI subjects. Our results thus demonstrate the applicability of the anti-saccade task as a cognitive marker in a-MCI. CONCLUSION: The work provides further support for eye-tracking as a useful diagnostic biomarker in the assessment of executive function in aging with cognitive impairments.

Behavioral and cognitive markers of mild cognitive impairment: diagnostic value of saccadic eye movements and Simon task.

BACKGROUND: Mild Cognitive Impairment (MCI) has been considered as a prodromal stage of Alzheimer disease (AD). Subtle changes in specific aspects of executive function like inhibitory control have been found in MCI. AIMS: We examined attentional and inhibitory control with the aim to distinguish between amnestic MCI patients and healthy controls. METHOD: Using neuropsychological, behavioral, and oculomotor function experiments, we examined executive function in 59 normal control, 49, multiple domain amnestic MCI (a-MCI) subjects, and 21 early stage AD patients using eye tracking and Simon task as measures of attentional control, to determine which saccade and behavioral tasks were sensitive enough to identify a-MCI. Saccades were investigated in gap and overlap pro-saccade and anti-saccade tasks. RESULTS: Scores on the Simon task were inversely correlated with general cognitive status and can distinguish a-MCI from controls with excellent specificity (AUC = 0.65 for reaction time and 0.59 for false responses). More importantly, our results showed that saccadic gains were affected in a-MCI and were the most sensitive measures to distinguish a-MCI from normal participants AST gap task AUC = 0.7, PST gap task AUC = 0.63, AST overlap task (AUC = 0.73). Moreover, these parameters were strongly correlated with neuropsychological measures. Using tests in parallel model, improved sensitivity up to 0.97. CONCLUSION: The present results enable us to suggest eye tracking along with behavioral data as a possible sensitive tools to detect a-MCI in preclinical stage.

Deep learning for diagnosis of head and neck cancers through radiographic data: a systematic review and meta-analysis.

PURPOSE: This study aims to review deep learning applications for detecting head and neck cancer (HNC) using magnetic resonance imaging (MRI) and radiographic data. METHODS: Through January 2023, a PubMed, Scopus, Embase, Google Scholar, IEEE, and arXiv search were carried out. The inclusion criteria were implementing head and neck medical images (computed tomography (CT), positron emission tomography (PET), MRI, Planar scans, and panoramic X-ray) of human subjects with segmentation, object detection, and classification deep learning models for head and neck cancers. The risk of bias was rated with the quality assessment of diagnostic accuracy studies (QUADAS-2) tool. For the meta-analysis diagnostic odds ratio (DOR) was calculated. Deeks' funnel plot was used to assess publication bias. MIDAS and Metandi packages were used to analyze diagnostic test accuracy in STATA. RESULTS: From 1967 studies, 32 were found eligible after the search and screening procedures. According to the QUADAS-2 tool, 7 included studies had a low risk of bias for all domains. According to the results of all included studies, the accuracy varied from 82.6 to 100%. Additionally, specificity ranged from 66.6 to 90.1%, sensitivity from 74 to 99.68%. Fourteen studies that provided sufficient data were included for meta-analysis. The pooled sensitivity was 90% (95% CI 0.820.94), and the pooled specificity was 92% (CI 95% 0.87-0.96). The DORs were 103 (27-251). Publication bias was not detected based on the p-value of 0.75 in the meta-analysis. CONCLUSION: With a head and neck screening deep learning model, detectable screening processes can be enhanced with high specificity and sensitivity.

Dichotic assessment of verbal memory function: development and validation of the Persian version of Dichotic Verbal Memory Test.

BACKGROUND AND PURPOSE: The Dichotic Verbal Memory Test (DVMT) is useful in detecting verbal memory deficits and differences in memory function between the brain hemispheres. The purpose of this study was to prepare the Persian version of DVMT, to obtain its results in 18- to 25-yr-old Iranian individuals, and to examine the ear, gender, and serial position effect. RESEARCH DESIGN: The Persian version of DVMT consisted of 18 10-word lists. After preparing the 18 lists, content validity was assessed by a panel of eight experts and the equivalency of the lists was evaluated. Then the words were recorded on CD in a dichotic mode such that 10 words were presented to one ear, with the same words reversed simultaneously presented to the other ear. Thereafter, it was performed on a sample of young, normal, Iranian individuals. STUDY SAMPLE: Thirty normal individuals (no history of neurological, ontological, or psychological diseases) with ages ranging from 18 to 25 yr were examined for evaluating the equivalency of the lists, and 110 subjects within the same age range participated in the final stage of the study to obtain the normative data on the developed test. RESULTS: There was no significant difference between the mean scores of the 18 developed lists (p > 0.05). The mean content validity index (CVI) score was .96. A significant difference was found between the mean score of the two ears (p < 0.05) and between female and male participants (p < 0.05). CONCLUSION: The Persian version of DVMT has good content validity and can be used for verbal memory assessment in Iranian young adults.