Machine learning methods in predicting the risk of malignant transformation of oral potentially malignant disorders: A systematic review

BACKGROUND

Oral Potentially Malignant Disorders (OPMDs) refer to a heterogenous group of clinical presentations with heightened rate of malignant transformation. Identification of risk levels in OPMDs is crucial to determine the need for active intervention in high-risk patients and routine follow-up in low-risk ones. Machine learning models has shown tremendous potential in several areas of dentistry that strongly suggest its application to estimate rate of malignant transformation of precancerous lesions.

METHODS

A comprehensive literature search was performed on Pubmed/MEDLINE, Web of Science, Scopus, Embase, Cochrane Library database to identify articles including machine learning models and algorithms to predict malignant transformation in OPMDs. Relevant bibliographic data, study characteristics, and outcomes were extracted for eligible studies. Quality of the included studies was assessed through the IJMEDI checklist.

RESULTS

Fifteen articles were found suitable for the review as per the PECOS criteria. Amongst all studies, highest sensitivity (100%) was recorded for U-net architecture, Peaks Random forest model, and Partial least squares discriminant analysis (PLSDA). Highest specificity (100%) was noted for PLSDA. Range of overall accuracy in risk prediction was between 95.4% and 74%.

CONCLUSION

Machine learning proved to be a viable tool in risk prediction, demonstrating heightened sensitivity, automation, and improved accuracy for predicting transformation of OPMDs. It presents an effective approach for incorporating multiple variables to monitor the progression of OPMDs and predict their malignant potential. However, its sensitivity to dataset characteristics necessitates the optimization of input parameters to maximize the efficiency of the classifiers.

Accuracy of machine learning in the diagnosis of odontogenic cysts and tumors: a systematic review and meta-analysis

BACKGROUND

The recent impact of artificial intelligence in diagnostic services has been enormous. Machine learning tools offer an innovative alternative to diagnose cysts and tumors radiographically that pose certain challenges due to the near similar presentation, anatomical variations, and superimposition. It is crucial that the performance of these models is evaluated for their clinical applicability in diagnosing cysts and tumors.

METHODS

A comprehensive literature search was carried out on eminent databases for published studies between January 2015 and December 2022. Studies utilizing machine learning models in the diagnosis of odontogenic cysts or tumors using Orthopantomograms (OPG) or Cone Beam Computed Tomographic images (CBCT) were included. QUADAS-2 tool was used for the assessment of the risk of bias and applicability concerns. Meta-analysis was performed for studies reporting sufficient performance metrics, separately for OPG and CBCT.

RESULTS

16 studies were included for qualitative synthesis including a total of 10,872 odontogenic cysts and tumors. The sensitivity and specificity of machine learning in diagnosing cysts and tumors through OPG were 0.83 (95% CI 0.81-0.85) and 0.82 (95% CI 0.81-0.83) respectively. Studies utilizing CBCT noted a sensitivity of 0.88 (95% CI 0.87-0.88) and specificity of 0.88 (95% CI 0.87-0.89). Highest classification accuracy was 100%, noted for Support Vector Machine classifier.

CONCLUSION

The results from the present review favoured machine learning models to be used as a clinical adjunct in the radiographic diagnosis of odontogenic cysts and tumors, provided they undergo robust training with a huge dataset. However, the arduous process, investment, and certain ethical concerns associated with the total dependence on technology must be taken into account. Standardized reporting of outcomes for diagnostic studies utilizing machine learning methods is recommended to ensure homogeneity in assessment criteria, facilitate comparison between different studies, and promote transparency in research findings.

Growth and digestive enzyme activities of rohu labeo rohita fed diets containing macrophytes and almond oil-cake

•

Rohu were fed with almond oil-cake/ duckweed/ water fern and fishmeal blend diets.

•

Highest growth was found in duckweed-based diet fed rohu.

•

Duckweed is a rich source of amino acids for rohu.

•

Feed composition influenced amylase, protease and lipase activities.

•

Duckweed helped to replace 300 g/kg dietary fishmeal without affecting growth.

The impact of plant-based diets on the digestive physiology of rohu Labeo rohita fingerlings (10.66 ± 0.53 g) was evaluated. A diet with all protein supplied by fishmeal was included as a control (F). Four test diets containing 300 g/kg protein were formulated using the following plant ingredients and fishmeal in a 1:1 blend: almond oil-cake Terminalia catappa (FTC), duckweed Lemna minor (FLM), water fern Salvania molesta (FSM) and combination of these three ingredients (FTCLMSM). The final body weight and specific growth rate were significantly higher in rohu fed diet FLM compared to the other treatments. Significantly lower feed conversion ratio in rohu fed diet FLM showed that diet was utilized efficiently in this feeding regime compared to the other diets. The composition of diets also influenced the digestive enzyme activities of the fish. Thus, amylase, trypsin and chymotrypsin activities were significantly higher in rohu fed diet FLM compared to the rohu fed the other diets. Protease activity was significantly higher in rohu fed diets FTC and F and lipase activity was significantly higher in rohu fed diet FTC compared to the rohu fed the other diets. The inclusion of raw duckweed in feed replaced 300 g/kg of dietary fishmeal without affecting growth.

Comparison of broiler chicken performance when fed diets containing meals of Bollgard II hybrid cotton containingCry-X gene(Cry1AcandCry2Ab gene), parental line or commercial cotton

(1) Total and free gossypol contents were 6.2 and 0.8, 5.4 and 0.5, and 6.1 and 0.7 g/kg in meals processed (solvent extracted) from Bollgard (BG) II, non-BG II or commercial cottonseeds, respectively. (2) Broiler chicks were given one of 7 dietary treatments (iso-nitrogenous, 220 and 195 g crude protein/ kg diet at 0 to 21 and 21 to 42 d, respectively, at a metabolisable energy concentration of 12.15 MJ/kg). The treatments were: D1 (control, soybean meal [SBM] based), D2 and D3 (commercial CSM at 100 g/kg of diet with and without additional iron), D4 and D5 (BG II CSM with and without additional iron), and D6 and D7 (non-BG II parental CSM with or without additional iron). (3) Body weight gain, feed intake, feed conversion efficiency, nutrient utilisation, certain blood constituents and carcase traits were not significantly affected by dietary treatments. (4) Weights of bursa and thymus were significantly higher in groups given diets containing BG II or non-BG diets containing added iron. (5) The results suggest that low free gossypol content cottonseed meals, for example, BG II, non-BG II and commercial solvent-extracted CSM could be included at 100 g/kg in broiler diets, safely replacing soybean meal without additional iron.

Calcium and phosphorus requirements of growing Japanese quail

1. Calcium and phosphorus requirements of growing female Japanese quail during 4 and 5 weeks of age were estimated in a 4 x 3 factorial experiment comprising 5, 7, 9 and 11 g calcium and 5, 6 and 7 g phosphorus/kg diet. 2. Responses in weight gain, bone ash, food intake and food : gain ratio showed no significant differences due to varying dietary calcium and phosphorus concentrations. 3. It appears that growing Japanese quail require not more than 5 g calcium and not more than 5 g phosphorus (2 g available phosphorus)/kg diet (398 mg calcium and 398 mg phosphorus or 159 mg available phosphorus MJ ME).