Criteria for assessing the diagnostic significance of modern methods of imaging gastrointestinal diseases in practical gastroenterology

Abstract

Imaging methods are frequently used to diagnose gastrointestinal diseases and play a crucial role in verifying clinical diagnoses among all diagnostic algorithms. However, these methods have limitations, challenges, benefits, and advantages. Addressing these limitations requires the application of objective criteria to assess the effectiveness of each diagnostic method. The diagnostic process is dynamic and requires a consistent algorithm, progressing from clinical subjective data, such as patient history (anamnesis), and objective findings to diagnostics ex juvantibus. Caution must be exercised when interpreting diagnostic results, and there is an urgent need for better diagnostic tests. In the absence of such tests, preliminary criteria and a diagnosis ex juvantibus must be relied upon. Diagnostic imaging methods are critical stages in the diagnostic workflow, with sensitivity, specificity, and accuracy serving as the primary criteria for evaluating clinical, laboratory, and instrumental symptoms. A comprehensive evaluation of all available diagnostic data guarantees an accurate diagnosis. The “gold standard” for diagnosis is typically established through either the results of a pathological autopsy or a lifetime diagnosis resulting from a thorough examination using all diagnostic methods.

Key Words: Imaging methods; Gastrointestinal diseases; Sensitivity; Specificity; Accuracy of the method

Core Tip: The diagnostic process is a complex journey that every physician undertakes with each patient. Successfully diagnosing gastrointestinal diseases requires mastery of all the methods within the diagnostic algorithm. Modern imaging methods provide physicians with significant diagnostic support. But how should the results of these imaging methods be evaluated? This is done using key criteria such as sensitivity, specificity, and accuracy. Only a comprehensive assessment of various diagnostic methods, taking into account these criteria, will ensure the correct diagnosis of the disease.

TO THE EDITOR

Imaging methods are frequently used to diagnose gastrointestinal diseases, serving as crucial tools to verify clinical diagnoses across various diagnostic algorithms (Table 1). These methods, however, have their limitations, challenges, benefits, and advantages. To address these limitations, it is necessary to apply objective criteria to evaluate the effectiveness of each diagnostic method. Non-invasive imaging techniques, such as ultrasound, CT, positron emission tomography (PET), and MRI, have revolutionized gastrointestinal diagnostics over the past few decades. Advancements in imaging resolution, three-dimensional imaging, and contrast agents have significantly improved diagnostic accuracy. Studies indicate remarkable diagnostic accuracy for various bowel conditions. For instance, inflammatory bowel diseases can be detected with 73%-87% sensitivity, while ulcerative colitis can be detected with 89% sensitivity and 100% specificity. Ultrasound also shows strong performance in diagnosing acute appendicitis (80%-93% sensitivity and 94%-100% specificity) and acute colonic diverticulitis (84%-100% sensitivity), achieving diagnostic accuracy comparable to that of CT scans[1]. In addition, literature reviews on the diagnosis of Crohn’s disease and its complications using small intestine contrast ultrasonography report sensitivity and specificity rates of 88% and 86%, respectively, for detecting small bowel lesions[2-4]. Assessing intestinal wall thickness further enhances the accuracy of this diagnostic method, with sensitivity, specificity, and accuracy values of 98%, 100%, and 98.3%, respectively[5]. Comparative studies reveal satisfactory performance for endoscopic studies and contrast-enhanced magnetic resonance (MR) enterography. Endoscopy has a sensitivity of 81.3% and a specificity of 70.5%, while MR enterography has a sensitivity of 80.2% and a specificity of 84.0%[6]. The diagnostic accuracy of imaging largely depends on the skill and expertise of the diagnostician. To ensure accurate interpretation and reduce diagnostic errors, radiologists need to thoroughly understand the factors that contribute to false-positive and false-negative findings[7,8]. To enhance diagnostic precision, optimal protocols tailored to the chosen imaging methods must be employed[9].

Table 1 Classification of diagnostic methods.

Groups of diagnostic methods	Criteria for assessing the diagnostic effectiveness of a method (symptom)
Clinical methods	Sensitivity, specificity, and accuracy
Subjective symptoms	-
Complaints	-
Anamnesis	-
Objective symptoms	-
Data of objective findings, somatic symptoms	-
Additional methods	-
Laboratory symptoms	-
Biochemical methods	-
Immuno-enzyme methods	-
Immunological methods	-
Molecular genetic methods	-
Bacteriological methods	-
Histopathological and cytological methods	-
Instrumental methods	-
Symptoms of radiation methods	-
Endoscopic symptoms	-
Symptoms of other methods	-
Method ex juvantibus	-

The reference “gold standard” method is typically established through either a pathological autopsy or a lifetime diagnosis obtained from a comprehensive examination that incorporates all available diagnostic methods.

IMAGING METHODS FOR DIAGNOSING DIGESTIVE ORGAN NEOPLASMS

Imaging techniques are essential for diagnosing gastrointestinal cancers, although their accuracy varies depending on the type of cancer and the method used. For gastric cancer, fasting whole-body PET/CT scans demonstrate a sensitivity of 92.9% and a specificity of 75%, with a positive predictive value of 94.5% and a negative predictive value of 69%. Enhancing these results by adding a mixture of milk and diatrizoate meglumine increases sensitivity to 91.1%, specificity to 91.7%, positive predictive value to 98.1%, and negative predictive value to 68.8%[10]. However, routine PET/CT scans may not be ideal for the initial staging of diffuse-type gastric cancer or for restaging lymph nodes after neoadjuvant treatment owing to lower sensitivities, which are reported at 24% and 32%, respectively. CT scans are useful in evaluating the primary gastric tumor and detecting liver metastasis, with sensitivity ranging from 54.5% to 72.7% and specificity from 89.3% to 94.6%, depending on the interpreting radiologist. The positive predictive value varies from 57.1% to 66.7%, while the negative predictive value ranges from 91.4% to 94.3%, highlighting the impact of radiologist interpretation on diagnostic accuracy[11,12]. For esophageal squamous cell carcinoma, multidetector CT shows variable diagnostic efficiency, with sensitivity ranging from 62.5% to 96.9%, specificity from 77.9% to 98.5%, and overall accuracy from 73% to 98%. The effectiveness of multidetector CT depends on the assessment criteria used, such as measuring the maximum esophageal wall thickness (≤ 9 mm) or the average attenuation of the esophageal wall (≤ 64 HU). These findings emphasize that, while imaging is indispensable for diagnosing and managing gastrointestinal cancers, factors such as the specific technique, the type and stage of cancer, the assessment criteria, and even the interpreting radiologist’s experience significantly influence the accuracy and reliability of the results[13]. There are conflicting results regarding the effectiveness of various imaging techniques for diagnosing gastrointestinal tumors[14-16]. However, diagnostic efficiency significantly improves when two modern imaging methods are combined, leading to substantially increased sensitivity, specificity, and accuracy[17-20]. High sensitivity, specificity, and accuracy of diagnostic methods not only enable the detection of disease but also help determine its activity and severity. For example, dual-energy CT enterography can measure iodine density, a criterion that reflects Crohn’s disease activity and correlates well with histological analysis[21]. This raises a question regarding the method that can serve as the reference standard when assessing the performance of imaging techniques (radiological or endoscopic diagnostics). Histopathological analysis is often seen as the gold standard, but it has its challenges. Various histopathological methods, such as biopsies in living patients, may yield conflicting results for the same disease in different individuals[22]. Factors influencing the accuracy of diagnostic methods include the technological sophistication of the equipment and the professional expertise of the diagnostician.

IMAGING METHODS FOR DIAGNOSING OTHER DISEASES

Imaging methods have high sensitivity and negative predictive value for diagnosing esophageal perforation. Thoracic CT has proven highly reliable in ruling out esophageal perforation, demonstrating 100% sensitivity and negative predictive value. This means that if the thoracic CT scan appears normal, patients can confidently be cleared of this complication. However, while the test excels at excluding perforation, it is less accurate in confirming its presence. Specifically, although the sensitivity for detecting esophageal perforation is a perfect 100%, ensuring that all perforations are identified, the specificity is lower at 54.6%, suggesting a higher chance of false positives. This is reflected in a positive predictive value of only 23.4%, meaning that only about one in four suspected cases based on the scan are true perforations[23]. For detecting choledocholithiasis, CT with contrast has moderate diagnostic effectiveness, with a sensitivity ranging from 77% to 88%, specificity ranging from 50% to 71%, and overall accuracy ranging from 71% to 74%[24]. MRI shows high diagnostic performance in pediatric appendicitis, with both sensitivity and specificity reaching 97%. Receiver operating characteristic analysis revealed an area under the curve of 0.98, indicating a high level of accuracy[25].

Diagnosing intestinal ischemia using clinical and laboratory methods is challenging. However, modern imaging methods offer improved diagnostic accuracy. For the first experienced radiologist, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy were 62.0%, 87.5%, 88.6%, 59.6%, and 72.0%, respectively, and for the second experienced radiologist, the corresponding values were 58.0%, 93.8%, 93.5%, 58.8%, and 72.0%[26]. To enhance the effectiveness of instrumental diagnostics, it is recommended to combine different imaging techniques and use multiple diagnostic approaches in conjunction.

CONCLUSION

The diagnostic process is dynamic and requires a consistent algorithm of diagnostic methods, from clinical subjective data (anamnesis) and objective findings to diagnostics ex juvantibus. Current diagnostic methods for these conditions are recognized as imperfect, necessitating caution in their application and underscoring the urgent need for more reliable diagnostic tools. Until such tools become available, clinicians must depend on preliminary criteria and diagnoses based on a patient’s response to treatment, which is inherently less reliable than a definitive diagnostic test[27,28]. Diagnostic imaging methods play a crucial role in this process. Sensitivity, specificity, and accuracy are key indicators used to evaluate the effectiveness of diagnostic methods across all diagnostic symptoms (clinical, laboratory, and instrumental). To avoid diagnostic errors, it is necessary to combine various instrumental diagnostic methods. The diagnostician must be highly trained, and it is recommended that two diagnosticians assess each imaging method. A comprehensive evaluation of all available diagnostic symptoms guarantees a correct diagnosis.