AI100 as a Screening Tool in Diagnosing Anemia - 3 Interesting Case Studies.

Renu Ethirajan¹, Arivarasan S P²

1. SigTuple Technologies, Bengaluru, Karnataka 560102, India.

2. Neuberg Bose Laboratories Pvt Ltd, Madurai, 62500, India.

*Correspondence to: Renu Ethirajan, SigTuple Technologies, Bengaluru, Karnataka 560102, India.

Copyright.

© 2025 Renu Ethirajan This is an open access article distributed under the Creative Commons Attribution   License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Received: 21 February 2025

Published: 06 March 2025

Abstract:

Background: One of the most common clinical conditions that we encounter regularly is anemia. The etiologies are vast and a careful approach to diagnosis will aid in the right management. Anemia may be due to decreased production of iron, vitamin B12 or folic acid or increased destruction as seen in hemolysis and hemoglobinopathies.⁽³⁾ Aplastic anemia is seen in conditions where the precursors in the marrow are absent or due to infiltration of the marrow by cancer cells as seen in leukemia, lymphoma or metastasis.⁽⁴⁾ Work up of anemias involves a thorough history of the patients, clinical examination, history of weight loss, nutrition, frequent transfusions, previous administration of iron, vitamin B12, or folic acid etc.

Artificial intelligence is increasingly being used in hematology to screen anemias and leukemias in peripheral blood smears, offering faster and more accurate analysis. In this article we discuss and share 3 interesting case reports how an AI based tool like AI100 powered by SHONIT aided the pathologist to give efficient and quick reports. The results of AI coupled with image based evidence and microscopic view provided on the reporting platform supported the pathologist in interpretation of the cases.

Conclusion: The use of AI in peripheral blood smear analysis aids in early detection, allowing quicker diagnosis and treatment. It supports hemato-pathologists to achieve a quick turn around time while increasing their productivity and efficiency. It significantly optimizes workflow and enhances patient care.  

Keywords: Anemia, AI assisted screening, Early Diagnosis, Healthcare Access, Artificial intelligence, AI100

Introduction

Anemia is a condition of red blood cell count below the normal threshold that serves as carrying oxygen to the body and remains a major public health problem worldwide, especially in developing countries.(1) It can result from nutritional deficiencies, chronic diseases, infections, and genetic disorders.(2) If the underlying cause of anemia is not diagnosed, it can lead to severe health issues such as fatigue, cognitive impairment, weakened immunity, and complications during pregnancy. Therefore, timely diagnosis is imperative to mitigate these risks.

The evaluation of anemia generally starts with a basic CBC test followed by a peripheral blood smear examination. There are various flags given by the CBC analyser which mandate the microscopic examination of a peripheral blood smear. Low Hb is one of them. The RBC morphology is a very important parameter which gives clues to the type of anemia.

During PBS examination the pathologist critically analyses the RBC morphology to see if they are microcytic or macrocytic. If macrocytic, then generally it is observed whether they are round or oval macrocytes. Round macrocytes are generally seen in conditions like hypothyroidism or alcoholic liver disease whereas oval macrocytes are seen in vitamin B12 and folic acid deficiencies. Schistocytes or fragmented RBCs are seen in hemolytic anemias.(3)

In vitamin B12 deficiencies and conditions like Myelodysplastic syndromes, there is ineffective erythropoiesis due to which the erythroblasts undergo apoptosis, resulting in anemia in the periphery. In patients, especially over 40 with a history of iron deficiency and weight loss, there may be an underlying gastro intestinal malignancy which needs to be ruled out.

In other cases one may have to rule out associated infections, hemolysis, bleeding etc. The presence of lymphadenopathy and hepatosplenomegaly is a very important finding which gives clues to underlying conditions like viral infections, lymphomas, acute leukemias, chronic myeloproliferative disorders, hairy cell leukemia, thalassemias etc.

The red cells of patients with aplastic anemia are normocytic or mildly macrocytic. Leukoerythroblastosis on the peripheral-blood smear suggests marrow replacement, vasculitis, or infection rather than aplastic anemia.(4) The investigation into the mechanism of anemia in conditions like leukemia, malignant lymphoma and myeloproliferative disorders has shown that a hemolytic process is frequently present.(7)

Keeping these points in view, we present 3 cases of anemia in different settings where the AI100 powered by SHONIT aided the pathologist in a quick analysis and interpretation.

Intended Purpose of the Device

AI100 powered by Shonit (Figure1) is an automated cell-locating device intended for in-vitro diagnostic use in clinical laboratories. AI100 with Shonit is intended to be used by operators trained in the use of the device. AI100 with Shonit is intended for differential count of White Blood Cells (WBC), characterization of Red Blood Cells (RBC) morphology and Platelet morphology.(6) It automatically locates blood cells on peripheral blood smears (PBS) and presents images of the blood cells for review. A skilled operator, trained in the use of the device and in the review of blood cells, identifies and classifies each cell according to type.

Figure 1:
 

Methods

The peripheral blood smear slides, prepared from the samples, were scanned by AI100 and the reports generated by the device were reviewed by the hemato-pathologist on the reporting platform (Mandara) at Neuberg Bose laboratory. (Fig1) and ( Fig2)

Figure 2
AI100 aided the pathologists while navigating through various cases of anemia. AI100 as a tool provides morphological analysis of RBCs, WBCs and platelets with visual evidence. The user-friendly microscopic view allows the pathologist to seamlessly navigate the smear and analyse the cells on the digital platform.

RBC analysis is simplified by AI, as it detects and classifies the RBCs according to size and shape while providing the grading according to the ICHS guidelines making it easy for the pathologist to interpret the RBC picture.(5) Nucleated RBCs are reported with high accuracy along with schistocytes in the fragmented RBC bucket which makes interpreting hemolytic anemias quick and easy. In addition to that, various poikilocytes, including target cells, elliptocytes, tear drop cells, and echinocytes, help distinguish the various categories of anemias in correlation with history, clinical and bio-chemical findings.

Case Reports

Shown below are a few cases of anemia reported on AI100

Case 1: Microcytic hypochromic blood picture.

AI100 flagged a significant presence of microcytes, grading them as 3+ based on the established guidelines, indicating a notable microcytic hypochromic blood picture.

Fig 3

Fig 4: Microscopic view of the microcytic hypochromic picture

Fig 5: Total WBC count and differentials were within normal limits.

The pathologist was able to give a diagnosis of microcytic hypochromic blood picture and recommend further tests for iron studies and Ferritin assays.

Case 2: Case of microcytic anemia (47Y / F)

AI100 flagged a significant presence of microcytes, grading them as 3+ based on the established guidelines. Fig 6: The smear also revealed fragmented cells and a few nRBCs.

Fig 7: The microscopic view showed few nRBCs and fragmented cells. Anisopoikilocytosis was seen.

The nRBCs and fragmented RBCs were flagged by AI100 along with visual evidence. This aided the pathologist to perform further tests to confirm early hemolysis.

Case 3. Anemia in a case of acute leukemia with thrombocytopenia.

Fig 8: AI100 showed 26.6% blasts on a background of microcytes

Fig 9: Microscopic view of the blasts on a background of microcytosis, and thrombocytopenia.

Please click here to view all images

This case was reported as Acute leukemia and further necessary evaluation such as immunophenotyping by flow cytometry was recommended

Conclusion

The pathologist was able to make quick decisions in all the above cases and proceed with further evaluations appropriate for each case. AI100 aided the pathologist by appropriately identifying and classifying the WBCs and RBCs while providing the visual evidence in each case, making it easier to interpret and proceed for further confirmatory tests. AI100 is an effective tool for screening anemia caused by various clinical conditions.

Digital pathology solutions have the potential to address the diagnostic, logistic and operational challenges in the laboratory environment. It can increase the throughput and decrease the turn around time; thereby increasing the productivity and efficiency of the pathologist. It creates an environment where a pathologist can gain remote access to digital images for verification and approval of reports. It obviates the need for the physical availability of slides and microscope during reporting, hence enabling remote access reporting. Manual error is markedly reduced by removing the ambiguity and bias associated with the human eye resulting in standardization, reproducibility and scalability.

References

1. S.R. Pasricha, K. Colman, E. Centeno-Tablante, et al.Revisiting WHO haemoglobin thresholds to define anaemia in clinical medicine and public healthLancet Haematol, 5 (2018), pp. E60-e62

2. American Society of Hematology. Anemia [Internet]; 2021. Available at: Anemia - Hematology.org

3. Hoffbrand AV, Catovsky D, Tuddenham EGD, Green AR (Eds.). Postgraduate Haematology, 6th edition.Oxford, United Kingdom: Wiley-Blackwell; 2011

4. Camitta, BM ? Storb, R ? Thomas, ED Aplastic anemia (part 2): pathogenesis, diagnosis, treatment, and prognosis N Engl J Med. 1982; 306:712-718

5. L. Palmer, C. Briggs, S. McFadden, G. Zini, J. Burthem, G. Rozenberg, M. Proytcheva, S. J. Machin ICSH recommendations for the standardization of nomenclature and grading of peripheral blood cell morphological feature First published: 02 March 2015

6. Upinder et al. Evaluation of automated blood microscopy system AI100 with Shonit in a tertiary care center in northern India IJIRMS, Vol 9, Issue 7 (2024)

7. Jane F.Desforges et al. Mechanisms of anemia in leukemia and malignant lymphoma. CLINICAL STUDY Volume 28, Issue 1 P69-76January 1960.