Below is the current list of 100 free radiology thesis topics for MD and DNB candidates in Radiodiagnosis. Each title uses a cross-sectional, observational, correlative or diagnostic accuracy design that a postgraduate can complete from patients already referred to the department for clinically indicated imaging, without prospective follow-up. Every topic generates a complete radiology protocol and radiology synopsis in editable format. For more topics you can avail the service of premium Radiology thesis topics.
Last reviewed and updated: August 2026
📌 Updated for 2026–2027 MD Radiodiagnosis admissions
This radiology thesis topic list is updated for the 2026–27 academic cycle. Topics are reviewed against recent dissertations, examiner preferences, feasibility in Indian medical colleges, and publication trends in radiodiagnosis.
Select Generate Protocol → beside any title and receive a submission-ready document built around that topic, containing all eighteen components:
Subscribing to the premium thesis topics not only lets you browse the full premium list, it also gives you online guidance (guidance doesnt mean complete protocol) on synopsis writing, sample size calculation, inclusion and exclusion criteria, and support throughout thesis writing. If you do not subscribe, please still feel free to contact me for guidance.
Unlock Premium Radiology Topics 🔒Studying outside India?
The topics above work as research questions anywhere — what changes is the document your institution expects. Two different routes, depending on which applies to you.
Board residents — SCFHS, Arab Board, OMSB, KIMS, QCHP, NHRA, DHA and DOH
Radiology residency across the Gulf carries a mandatory research requirement, and the equivalent of an Indian synopsis is the research proposal submitted to your IRB before a research project begins. The format differs from the Indian one: it additionally requires a Gantt chart, a budget and resources section, and a Declaration of Helsinki statement.
Generate a residency research proposal →PhD and Master's candidates — Saudi Arabia, Malaysia, the Gulf and beyond
University graduate programmes generally require a full research proposal of roughly 6,000 to 10,000 words, with an extended literature review, a theoretical framework and a detailed methodology chapter — considerably longer and deeper than a residency proposal. These are written individually, by a medical doctor, with no artificial intelligence generation and no plagiarism, and revised until your supervisor accepts them.
Enquire about a PhD research proposal →🔥 Trending research areas in Radiology for 2026–27
Based on recent thesis submissions and examiner preferences in MD Radiodiagnosis departments across India, these are the emerging high-interest areas:
A radiology protocol is judged on internal consistency: the research question, objectives, methodology and statistical plan must all describe the same study. The primary objective should be a single measurable endpoint — one diagnostic comparison, one correlation, one imaging profile — with everything else demoted to secondary objectives.
Equipment and acquisition must be specified fully, because a finding is meaningless without the technique that produced it. State the machine make and model; for magnetic resonance imaging the field strength, coil and the sequences with their parameters; for computed tomography the number of detector rows, slice thickness, reconstruction algorithm and whether contrast was used, with the agent, volume, injection rate and phase timing; for ultrasonography the transducer type and frequency and the Doppler settings. Where measurements are made, say on which plane, at which anatomical landmark, and how many times.
Name the reporting or classification system with its version: BI-RADS for breast, TI-RADS for thyroid, LI-RADS where liver lesions are categorised, ASPECTS for early ischaemic change, the modified computed tomography severity index for pancreatitis, the Fleischner recommendations for pulmonary nodules, Wells or another named clinical probability score for pulmonary embolism, Parks classification with the St James's University Hospital grading for perianal fistula, and the grading system used for hydronephrosis and vesicoureteral reflux.
Two elements decide whether a diagnostic study is credible. First, the reference standard: state explicitly what establishes the truth — histopathology, cytology, operative findings, or a defined clinical course — and confirm it is applied to all participants, not only to those the imaging called abnormal. Second, blinding and observer arrangements: the radiologist interpreting the images must be blinded to the reference standard and to the clinical details beyond the indication, at least two readers should interpret a defined subset so that interobserver agreement can be reported, and the protocol must say how disagreement is resolved. State the readers' level of experience, since that materially affects accuracy.
A radiology synopsis is the condensed document of two to four pages — title, introduction, aim and objectives, brief methodology, sample size and references — submitted for registration of the dissertation topic. The radiology protocol is the expanded version of twelve to twenty pages carrying the full review of literature, detailed methodology including equipment, acquisition parameters, reporting systems, reference standard and observer arrangements, statistical plan, study timeline and annexures including the patient information sheet, consent form and image evaluation proforma.
In practice the synopsis is extracted from the protocol rather than written separately, which is faster and produces a more coherent document. Check your university's prescribed proforma before submission, since rejections on formatting grounds are common and entirely avoidable.
Most of this list consists of diagnostic accuracy studies, and they require a calculation based on the expected sensitivity or specificity, the precision with which you want to estimate it, and the expected prevalence of disease among those imaged. Using a plain proportion formula here is the commonest sample size error in radiology protocols: only the diseased participants contribute to the sensitivity estimate, so a study of two hundred patients in which twenty have the condition is estimating sensitivity from twenty cases. Show that arithmetic. Imaging profile and prevalence studies use a proportion-based calculation; correlation studies between an imaging measurement and a clinical or laboratory parameter use the expected correlation coefficient; agreement studies use a calculation based on the expected kappa.
Name the tests rather than promising that data will be analysed. Diagnostic accuracy studies report sensitivity, specificity, positive and negative predictive values and diagnostic accuracy with confidence intervals against the stated reference standard, and where a continuous measurement is being evaluated for a threshold, receiver operating characteristic analysis with area under the curve and a derived cut-off. Agreement between two modalities or two readers is reported as a kappa statistic for categorical findings and by Bland-Altman analysis for continuous measurements, not as a correlation coefficient — two measurements can correlate perfectly and still disagree systematically, which is a standard criticism of imaging comparisons. Comparison of proportions uses the chi-squared test with Fisher's exact test for small cells, and comparison of a continuous measure between groups uses the independent t-test or the Mann-Whitney U test with a stated normality test.
Where both sides of a paired organ or multiple lesions in one patient are assessed, state whether the unit of analysis is the patient or the lesion, and account for clustering rather than treating lesions from the same patient as independent.
Start with your department's workload register. Count how many studies of the relevant type were performed over the last twelve months, since referral volume rather than patient numbers determines what you can collect. Ultrasonography, computed tomography of the abdomen and chest, and antenatal scanning accrue quickly. Magnetic resonance imaging of the breast, computed tomography pulmonary angiography and interventional procedures are far less frequent in most teaching departments.
Then check whether the reference standard will actually be available. A study correlating imaging with histopathology only works if those patients are reliably biopsied at your institution and the reports are traceable — many are referred elsewhere for surgery and never return. Confirm this with the surgical and pathology departments before committing, because it is the most common reason radiology dissertations stall at analysis.
Cross-sectional diagnostic accuracy studies dominate: one modality against another, imaging against histopathology, or a structured reporting category against the eventual diagnosis. Imaging profile and spectrum studies of a defined condition, correlation studies between an imaging measurement and a clinical, functional or laboratory parameter, and prevalence studies of a finding within a routinely imaged population are all equally well established.
Retrospective designs using archived studies are accepted and are often the only way to accumulate enough cases of an uncommon condition. Interventional and follow-up designs are rare at postgraduate level.
Bring three to five shortlisted titles rather than one, since guides frequently rule out a topic on grounds you could not have known — an ongoing departmental study, a machine due for replacement or extended servicing, a referral stream that has changed.
Settle four things in that meeting: annual study volume in the relevant modality, whether the reference standard is reliably obtainable and from which department, who will act as the second independent reader for agreement analysis, and which journal the eventual paper is aimed at.
Yes, and for uncommon conditions it is often the only viable route. State the archive period searched, how cases were identified from the reporting system, and the completeness of the retrieved data. Two cautions. Acquisition parameters may have varied over the period, so record them per case rather than describing a single protocol. And where reports rather than images are the data source, say whether images were re-read for the study or the original report accepted as recorded — examiners will ask, and a study built on original reports cannot claim blinded assessment.
The synopsis is the condensed two to four page document submitted for topic registration. The protocol is the full document of twelve to twenty pages containing the detailed review of literature, methodology with equipment, acquisition parameters, reporting systems, reference standard and observer arrangements, statistical plan, timeline and annexures. The synopsis is normally extracted from the completed protocol.
Full institutional ethics committee approval before data collection. Three points need explicit attention in this speciality.
Radiation. State clearly whether the imaging is clinically indicated and would have been performed anyway, or whether any examination is being done for research. Committees will not readily approve additional ionising radiation for a dissertation, and a protocol that leaves this ambiguous invites questions. Where children are imaged, address it directly.
Contrast. Where contrast-enhanced studies are included, state the screening for renal impairment and previous reactions, and that administration follows departmental protocol for the clinical indication rather than for the study.
Anonymisation. Retrospective image-based studies need particular care because patient identifiers are embedded in the image metadata itself, not only in the report. State that studies are de-identified before analysis, where the anonymised data are stored, and that no identifying information appears in any image reproduced in the dissertation or a publication.
Purely retrospective studies may be granted a waiver of consent, applied for explicitly. Prospective studies on clinically indicated imaging still need written informed consent for participation and data use. Clearance commonly takes six to ten weeks and retrospective approval is not granted.
Name it, and apply it to everyone. The commonest flaw in radiology dissertations is that the reference standard is obtained only for patients whose imaging was abnormal: if only the suspicious lesions go to biopsy, sensitivity is overestimated and specificity cannot be assessed at all, because the imaging result has determined who gets verified. If universal verification is impossible — and it usually is — state how the remaining patients are classified, whether by a defined clinical course over a stated interval or by a second imaging modality, and acknowledge the limitation in the discussion plan. Also state the interval between the index imaging and the reference standard, since a long gap allows the disease itself to change. Following a recognised reporting guideline for diagnostic accuracy studies when writing up makes all of this straightforward, and reviewers increasingly expect it.
Substantially. A PhD proposal typically runs 6,000 to 10,000 words and carries an extended critical literature review, a theoretical framework, a detailed methodology chapter and a discussion of expected contribution to the field. An MD synopsis is a two to four page registration document. The research question can be the same; the depth expected is not.
Most universities require registration within six to nine months of joining. Shortlist in the first two months, finalise with your guide by the third, and file for ethics clearance immediately afterwards. Where pathology, surgery or another clinical department must supply your reference standard, secure that agreement in writing at the same time rather than assuming informal cooperation will hold for two years.
Found a topic that fits your department?
Generate its full protocol — objectives, methodology, sample size, statistics, timeline, references and annexures — in editable format.
Generate your protocol
