Role of artificial intelligence, machine learning and deep learning models in corneal disorders - A narrative review

In the last decade, artificial intelligence (AI) has significantly impacted ophthalmology, particularly in managing corneal diseases, a major reversible cause of blindness. This review explores AI's transformative role in the corneal subspecialty, which has adopted advanced technology for superior clinical judgment, early diagnosis, and personalized therapy. While AI's role in anterior segment diseases is less documented compared to glaucoma and retinal pathologies, this review highlights its integration into corneal diagnostics through imaging techniques like slit-lamp biomicroscopy, anterior segment optical coherence tomography (AS-OCT), and in vivo confocal biomicroscopy. AI has been pivotal in refining decision-making and prognosis for conditions such as keratoconus, infectious keratitis, and dystrophies. Multi-disease deep learning neural networks (MDDNs) have shown diagnostic ability in classifying corneal diseases using AS-OCT images, achieving notable metrics like an AUC of 0.910. AI's progress over two decades has significantly improved the accuracy of diagnosing conditions like keratoconus and microbial keratitis. For instance, AI has achieved a 90.7% accuracy rate in classifying bacterial and fungal keratitis and an AUC of 0.910 in differentiating various corneal diseases. Convolutional neural networks (CNNs) have enhanced the analysis of color-coded corneal maps, yielding up to 99.3% diagnostic accuracy for keratoconus. Deep learning algorithms have also shown robust performance in detecting fungal hyphae on in vivo confocal microscopy, with precise quantification of hyphal density. AI models combining tomography scans and visual acuity have demonstrated up to 97% accuracy in keratoconus staging according to the Amsler-Krumeich classification. However, the review acknowledges the limitations of current AI models, including their reliance on binary classification, which may not capture the complexity of real-world clinical presentations with multiple coexisting disorders. Challenges also include dependency on data quality, diverse imaging protocols, and integrating multimodal images for a generalized AI diagnosis. The need for interpretability in AI models is emphasized to foster trust and applicability in clinical settings. Looking ahead, AI has the potential to unravel the intricate mechanisms behind corneal pathologies, reduce healthcare's carbon footprint, and revolutionize diagnostic and management paradigms. Ethical and regulatory considerations will accompany AI's clinical adoption, marking an era where AI not only assists but augments ophthalmic care.

Effect of microwave drying of paddy on milling characteristics, cooking quality, and microstructure of rice

Effect of microwave drying on milling, cooking, and microstructure characteristic of paddy rice in comparison to hot air drying was assessed in the present study. Results revealed that modified microwave drying significantly (P < 0.05) affected the milling quality of paddy. Raw paddy had head rice recovery of 49.63%, while after microwave drying, the head rice recovery increased by 6.73% in comparison to hot air drying. Microwave drying brought significant changes in the colour characteristic of rice as it had total colour change of 13.50 in comparison to 10.93 by hot air drying. Cooking time and water uptake ratio after microwave drying increased to 31.46 min and 3.16%, in comparison to 27.05 min and 2.65% for hot air dried samples, respectively. Scanning Electron Microscope images revealed that both hot air and microwave dried rice had a coarse surface and large starch particles, while starch structure was more damaged in hot air drying, however, agglomeration of protein-starch matrix was more uniform in microwave dried samples due to modification of the microwave applicator that prevented thermal decomposition. It was recommended that paddy may be dried using a modified microwave applicator with a shorter duration and a better quality.

A retrospective audit of endoscopic duodenal biopsies to uncover undetected Coeliac disease in Malaysian patients

BACKGROUND

Coeliac disease, an autoimmune enteropathy related to gluten sensitivity was hitherto thought to be rare in Asia. Recent data however suggests that Celiac disease may be under-diagnosed in Asia.

OBJECTIVE

The aim of this audit was to determine the frequency of histological changes compatible with Coeliac disease among patients undergoing elective diagnostic oesaphago-gastro-duodenoscopy (OGDS) under the care of a single practitioner in a Malaysian hospital.

MATERIALS AND METHODS

The archived endoscopically obtained duodenal biopsy specimens of 241 consecutive Malaysian subjects undergoing elective diagnostic (OGDS) were reviewed by a pathologist blinded to the clinical data. Based on intra-epithelial lymphocyte counts, crypt hyperplasia and villous atrophy, each subject was assigned to one of the categories of the Modified Marsh classification for the histological diagnosis of Coeliac disease. The clinical charts of all subjects were reviewed by a single gastroenterologist blinded to the findings of the histological review.

RESULTS

Of the 241 study subjects, 132 (54.8%) were females. There were 56 (23.2%) Malays, 90 (37.3%) Chinese, 88 (36.5%) Indians and seven (2.9%) from the other category. The median age of the study sample was 49 years (range 15- 88 years). The OGDS was done as part of screening in 15(6.2%) subjects while in the remaining it was part of the investigation of a clinical problem. Based on histological findings, none of the subjects could be assigned to a modified Marsh class of >1. The prevalence of histological changes compatible with Coeliac disease in the study was 0% (binomial exact one-sided 97.5 % confidence interval 0- 1.52%).

CONCLUSION

In conclusion, this audit provides no evidence that active Coeliac disease is significantly under-detected among symptomatic patients presenting for diagnostic OGDS. The possibility that a significant number may have potential coeliac disease cannot be excluded.

Impact of ophthalmic webinars on the resident's learning experience during COVID-19 pandemic: An insight into its present and future prospects

Purpose:

To analyze the impact of ophthalmic webinars on the resident's learning experience during the COVID-19 pandemic (CP).

Methods:

This cross-sectional nationwide study was carried out for 1 month during CP and included a total of 382 ophthalmic residents. A questionnaire was sent through various social media platforms.

Results:

Residents expressed a decline in their clinical exposure (74%; 220), thesis work (58%; 218), and acquisition of the knowledge and skills (42.5%; 161) during CP. Benefits of webinars as perceived by the residents included gain in additional knowledge (77%; 286), feedback on queries (56%; 209), access to multiple speakers (50%; 191), and topics (30%; 110). Nearly 75% (291) of residents endorsed webinars as good to the very good academic tool, and 54% (202) preferred to continue attending webinars in the post-CP phase. However, connectivity/download/data issues (54%; 200) followed by loss of personal touch (53%; 188), lengthy or irrelevant topic (37%; 134), and poor transmission quality (33%; 121) were major deterrents against the webinar.

Conclusion:

The current study generated overall mixed responses from the ophthalmic postgraduate residents in favor of webinars. In the present format, webinars bear enormous potentials to supplement the traditional learning tools by providing uninterrupted learning experiences. However, they are still limited by their pedagogical and technical issues.

Role of magnetic resonance imaging in breast cancer management

Magnetic resonance imaging (MRI) of the breast is primarily used as a supplemental tool to breast screening with mammography or ultrasound. A breast MRI is mainly used for women who have been diagnosed with breast cancer, to help measure the size of the cancer, look for other tumors in the breast, and to check for tumors in the opposite breast. For certain women at high risk for breast cancer, a screening MRI is recommended along with a yearly mammogram. MRI is known to give some false positive results which mean more test and/or biopsies for the patient. Thus, although breast MRI is useful for women at high risk, it is rarely recommended as a screening test for women at average risk of breast cancer. Also, breast MRI does not show calcium deposits, known as micro-calcifications which can be a sign of breast cancer.

Multicopper oxidase of Acinetobacter baumannii: Assessing its role in metal homeostasis, stress management and virulence

A putative multicopper oxidase, encoded as CopA in the proteome of Acinetobacter baumannii 19606, and designated as AbMCO, was expressed heterologously in E. coli (pET-28a) and purified by Ni-NTA affinity chromatography. The purified AbMCO exhibited in vitro oxidase activities upon exogenous addition of ≥1 μM copper ions. Kinetic studies revealed its phenol oxidase activity as it could catalyze the oxidation of substrates viz. 2,2'-azino-bis(3-ethylbenzothiazoline-6-sulphonic acid) (ABTS), guaiacol, pyrogallol and catechol. Additionally, AbMCO displayed siderophore oxidase activity which depicted its role in metal homeostasis and protection from the toxic redox states of copper and iron. Importantly, expression of abMCO increased manifold upon challenge with high concentrations of copper sulphate (CuSO₄, 1.5 mM) and sodium chloride (NaCl, 700 mM) which suggested its protective role in stress adaptation and management. Intra-macrophage assay of abMCO-expressing and abMCO-non expressing cells depicted no significant change in the survival rate of A. baumannii inside the macrophages. These findings indicate that A. baumannii encodes a multicopper oxidase, conferring copper tolerance and survival under stress conditions but had no role in virulence of this pathogen.

Development of a calibration chamber to evaluate the performance of low-cost particulate matter sensors

Low-cost particulate matter (PM) air quality sensors are becoming widely available and are being increasingly deployed in ambient and home/workplace environments due to their low cost, compactness, and ability to provide more highly resolved spatiotemporal PM concentrations. However, the PM data from these sensors are often of questionable quality, and the sensors need to be characterized individually for the environmental conditions under which they will be making measurements. In this study, we designed and assessed a cost-effective (∼$700) calibration chamber capable of continuously providing a uniform PM concentration simultaneously to multiple low-cost PM sensors and robust calibration relationships that are independent of sensor position. The chamber was designed and evaluated with a Computational Fluid Dynamics (CFD) model and a rigorous experimental protocol. We then used this new chamber to calibrate 242 Plantower PMS 3003 sensors from two production lots (Batches I and II) with two aerosol types: ammonium nitrate (for Batches I and II) and alumina oxide (for Batch I). Our CFD models and experiments demonstrated that the chamber is capable of providing uniform PM concentration to 8 PM sensors at once within 6% error and with excellent reliability (intraclass correlation coefficient > 0.771). The study identified two malfunctioning sensors and showed that the remaining sensors had high linear correlations with a DustTrak monitor that was calibrated for each aerosol type (R2 > 0.978). Finally, the results revealed statistically significant differences between the responses of Batches I and II sensors to the same aerosol (P-value<0.001) and the Batch I sensors to the two different aerosol types (P-value<0.001). This chamber design and evaluation protocol can provide a useful tool for those interested in systematic laboratory characterization of low-cost PM sensors.

P04.19 Analysis of tumor immune microenvironment and immune checkpoint modulators across infantile and pediatric pilocytic astrocytomas to elucidate the role of immunotherapy in these tumors

BACKGROUND

Pilocytic astrocytomas are the most common central nervous system tumors in pediatric age-group. Although grade I, some of the cases show recurrence and progression, and few might not be amenable to surgery due to location or size, and hence have a less favorable prognosis. Drugs blocking immune check-point interactions such as those including programmed cell death ligand-1 (PD-L1) and cytotoxic T-lymphocyte-associated protein 4 (CTLA-4) are now in clinical use for certain tumors. We performed this study to understand the potential candidature of pilocytic astrocytomas in infants and children for immunotherapy by analyzing the expression of immune checkpoint proteins and immune infiltrate, and correlating with clinical details, wherever possible.

MATERIALS AND METHODS

Cases with adequate tissue (2010–2017) diagnosed in pediatric age-group (<18 years) were retreived from the archives of Department of Pathology, AIIMS, New Delhi. Immunohistochemistry for PD-L1 (SP263, Ventana), CTLA-4, CD3, CD8, CD4 and CD68 was performed. Quantification of cytotoxic lymphocytes was done using digital imaging in the core of the tumor.

RESULTS

A total of 50 pilocytic astrocytomas were included, 14 of them were <3 years (infants), while 36 were of pediatric age-group (3–18 years). Overall, male preponderance was noted. Cerebellum was the most common location, followed by 4th venrticle, optic pathway, hypothalamus, cerebrum and thalamus. Almost all CD3 lymphocytes were cytotoxic T-lymphocyes (CD8 positive, CTLs). Helper T-lymphocyte infiltration was not seen. Median CTL density/mm3 was 13/mm3(Range:1–85/mm3). CTLA-4 was positive in 4 cases, positivity ranged from 1–4 cells/lpf. PD-L1 was found to be positive in 7 cases, and the positivity ranged from 1+ to 2+ in 1 to 5% of tumor cells. A median TAM (tumor associated macrophages) density of 44/hpf (range: 1–98/hpf) was noted. There was no correlation of CTL density with PD-L1 or CTLA-4 expression, and neither with TAM density. On correlation with clinical parameters, a higher density of CTLs and TAMs was noted in infants, and a higher proportion of cases revealed PD-L1 positivity, though not statistically significant. There was no correlation of TILs or TAMs with the tumor location.

CONCLUSION

Immune check point blockade using PD-(L)1 or CTLA4 inhibitors may not be a potential therapeutic option for unresectable or recurrent pilocytic astrocytomas, as low positivity rate as well as extremely low percentage of tumor/ immune cells found to be positive. However, alternate forms of immunotherapy might be helpful as most of the cases showed immune infiltrates and a high density of tumor-associated macrophages (TAMs). Large scale studies with larger numbers and longer follow-up periods including in-vitro and clinical studies are warranted for decoding the tumor immunogram.

P14.117 Cost efficient test algorithm for molecular subgrouping of medulloblastomas for day-to-day practice in resource limited countries

INTRODUCTION

Molecular classification of medulloblastomas(MB) is prognostically and therapeutically relevant and helps in better risk-stratification. Translation of this subgrouping to routine practice still remains a challenge. The most pathologist-accessible techniques for molecular subgrouping include immunohistochemistry(IHC), fluorescent in-situ hybridization(FISH) and Nanostring assay. Objectives:(1)Molecular subgrouping of MBs by IHC and FISH, and Nanostring Assay (2)To compare their efficacy against sequencing and DNA methylation, and cost for applicability in resource-constrained centers

METHODS

Ninety-five cases of MB with adequate tissue were included. Molecular subgrouping was performed by IHC for β-catenin, GAB1, YAP1, and p53; FISH for MYC amplification, and sequencing for CTNNB1, and by Nanostring Assay on the same set of MBs. Further, a subset of cases were subjected to 850k DNA methylation array.

RESULTS

IHC+FISH classified MBs into 15.8% WNT, 16.8% SHH, and 67.4% non-WNT/non-SHH subgroups; with MYC amplification identified in 20.3% cases of non-WNT/non-SHH. A single showed diffuse strong p53 positivity among the SHH subgroup. Nanostring successfully classified 91.5% MBs into 25.3% WNT, 17.2% SHH, 23% Group3 and 34.5% Group4. However, Nanostring assay failure was seen in eight cases, all of which were >8-years-old formalin-fixed paraffin-embedded tissue blocks. Concordant subgroup assignment was noted in 88.5% cases, while subgroup switching was seen in 11.5% cases. Both methods showed prognostic correlation. Among the 5 discrepant cases, which switched to WNT subgroup by Nanostring, only 2 were found to have CTNNB1 mutation. Methylation profiling performed on discordant cases revealed 1 out of 4 extra WNT identified by Nanostring to be WNT, others aligned with IHC subgroups; extra SHH by Nanostring turned out to be SHH by methylation.

CONCLUSIONS

Both IHC supplemented by FISH and Nanostring are robust methods for molecular subgrouping, albeit with few disadvantages. IHC cannot differentiate between Groups 3 and 4, while Nanostring cannot classify older-archived tumors, and is not available at most centres. WNT subgroup with the best prognosis is best detected by IHC or IHC followed by sequencing for confirmation. Nanostring Assay is better suitable to separate Group 3, the worst prognostic group from Group 4. Thus, both the methods complement each other and can be used in concert for high confidence allotment of molecular subgroups in clinical practice. The cost of IHC plus Nanostring will almost be the same as IHC plus FISH. We recommend a cost-efficient algorithmic approach using histopathological subtype and IHC as the first step followed by Nanostring or FISH, wherever suitable.

Engineered Peptides for Applications in Cancer-Targeted Drug Delivery and Tumor Detection

Cancer-targeting peptides as ligands for targeted delivery of anticancer drugs or drug carriers have the potential to significantly enhance the selectivity and the therapeutic benefit of current chemotherapeutic agents. Identification of tumor-specific biomarkers like integrins, aminopeptidase N, and epidermal growth factor receptor as well as the popularity of phage display techniques along with synthetic combinatorial methods used for peptide design and structure optimization have fueled the advancement and application of peptide ligands for targeted drug delivery and tumor detection in cancer treatment, detection and guided therapy. Although considerable preclinical data have shown remarkable success in the use of tumor targeting peptides, peptides generally suffer from poor pharmacokinetics, enzymatic instability, and weak receptor affinity, and they need further structural modification before successful translation to clinics is possible. The current review gives an overview of the different engineering strategies that have been developed for peptide structure optimization to confer selectivity and stability. We also provide an update on the methods used for peptide ligand identification, and peptide- receptor interactions. Additionally, some applications for the use of peptides in targeted delivery of chemotherapeutics and diagnostics over the past 5 years are summarized.