Evaluating Temperature Scaling Calibration Effectiveness for CNNs under Varying Noise Levels in Brain Tumour Detection Article Swipe

PDF

Ankush Chanda , Kaushik Roy Choudhury , Sanjoy Roy , Suman Biswas , Somenath Kuiry ·

YOU? · · 2025 · Open Access · · DOI: https://doi.org/10.48550/arxiv.2509.24951

Precise confidence estimation in deep learning is vital for high-stakes fields like medical imaging, where overconfident misclassifications can have serious consequences. This work evaluates the effectiveness of Temperature Scaling (TS), a post-hoc calibration technique, in improving the reliability of convolutional neural networks (CNNs) for brain tumor classification. We develop a custom CNN and train it on a merged brain MRI dataset. To simulate real-world uncertainty, five types of image noise are introduced: Gaussian, Poisson, Salt & Pepper, Speckle, and Uniform. Model performance is evaluated using precision, recall, F1-score, accuracy, negative log-likelihood (NLL), and expected calibration error (ECE), both before and after calibration. Results demonstrate that TS significantly reduces ECE and NLL under all noise conditions without degrading classification accuracy. This underscores TS as an effective and computationally efficient approach to enhance decision confidence of medical AI systems, hence making model outputs more reliable in noisy or uncertain settings.

Related Topics

Temperature

Wet-Bulb Temperature

U.S. State And Territory Temperature Extremes

Room Temperature

List Of Countries By Average Yearly Temperature

Lowest Temperature Recorded On Earth

Color Temperature

Wet-Bulb Globe Temperature

Temperature (Song)

Deborah Scaling Kiley

Feature Scaling

Ultra-High-Temperature Processing

Concepts

No concepts available.

Metadata

Type: preprint
Language: en
Landing Page: http://arxiv.org/abs/2509.24951
PDF: https://arxiv.org/pdf/2509.24951
OA Status: green
OpenAlex ID: https://openalex.org/W4415337446

All OpenAlex metadata

Raw OpenAlex JSON

OpenAlex ID: https://openalex.org/W4415337446

Canonical identifier for this work in OpenAlex
DOI: https://doi.org/10.48550/arxiv.2509.24951

Digital Object Identifier
Title: Evaluating Temperature Scaling Calibration Effectiveness for CNNs under Varying Noise Levels in Brain Tumour Detection

Work title
Type: preprint

OpenAlex work type
Language: en

Primary language
Publication year: 2025

Year of publication
Publication date: 2025-09-29

Full publication date if available
Authors: Ankush Chanda, Kaushik Roy Choudhury, Sanjoy Roy, Suman Biswas, Somenath Kuiry

List of authors in order
Landing page: https://arxiv.org/abs/2509.24951

Publisher landing page
PDF URL: https://arxiv.org/pdf/2509.24951

Direct link to full text PDF
Open access: Yes

Whether a free full text is available
OA status: green

Open access status per OpenAlex
OA URL: https://arxiv.org/pdf/2509.24951

Direct OA link when available
Cited by: 0

Total citation count in OpenAlex

Full payload

id	https://openalex.org/W4415337446
doi	https://doi.org/10.48550/arxiv.2509.24951
ids.doi	https://doi.org/10.48550/arxiv.2509.24951
ids.openalex	https://openalex.org/W4415337446
fwci
type	preprint
title	Evaluating Temperature Scaling Calibration Effectiveness for CNNs under Varying Noise Levels in Brain Tumour Detection
biblio.issue
biblio.volume
biblio.last_page
biblio.first_page
topics[0].id	https://openalex.org/T12994
topics[0].field.id	https://openalex.org/fields/27
topics[0].field.display_name	Medicine
topics[0].score	0.9864000082015991
topics[0].domain.id	https://openalex.org/domains/4
topics[0].domain.display_name	Health Sciences
topics[0].subfield.id	https://openalex.org/subfields/2741
topics[0].subfield.display_name	Radiology, Nuclear Medicine and Imaging
topics[0].display_name	Infrared Thermography in Medicine
topics[1].id	https://openalex.org/T11948
topics[1].field.id	https://openalex.org/fields/25
topics[1].field.display_name	Materials Science
topics[1].score	0.9595999717712402
topics[1].domain.id	https://openalex.org/domains/3
topics[1].domain.display_name	Physical Sciences
topics[1].subfield.id	https://openalex.org/subfields/2505
topics[1].subfield.display_name	Materials Chemistry
topics[1].display_name	Machine Learning in Materials Science
topics[2].id	https://openalex.org/T10320
topics[2].field.id	https://openalex.org/fields/17
topics[2].field.display_name	Computer Science
topics[2].score	0.9320999979972839
topics[2].domain.id	https://openalex.org/domains/3
topics[2].domain.display_name	Physical Sciences
topics[2].subfield.id	https://openalex.org/subfields/1702
topics[2].subfield.display_name	Artificial Intelligence
topics[2].display_name	Neural Networks and Applications
is_xpac	False
apc_list
apc_paid
language	en
locations[0].id	pmh:oai:arXiv.org:2509.24951
locations[0].is_oa	True
locations[0].source.id	https://openalex.org/S4306400194
locations[0].source.issn
locations[0].source.type	repository
locations[0].source.is_oa	True
locations[0].source.issn_l
locations[0].source.is_core	False
locations[0].source.is_in_doaj	False
locations[0].source.display_name	arXiv (Cornell University)
locations[0].source.host_organization	https://openalex.org/I205783295
locations[0].source.host_organization_name	Cornell University
locations[0].source.host_organization_lineage	https://openalex.org/I205783295
locations[0].license
locations[0].pdf_url	https://arxiv.org/pdf/2509.24951
locations[0].version	submittedVersion
locations[0].raw_type	text
locations[0].license_id
locations[0].is_accepted	False
locations[0].is_published	False
locations[0].raw_source_name
locations[0].landing_page_url	http://arxiv.org/abs/2509.24951
locations[1].id	doi:10.48550/arxiv.2509.24951
locations[1].is_oa	True
locations[1].source.id	https://openalex.org/S4306400194
locations[1].source.issn
locations[1].source.type	repository
locations[1].source.is_oa	True
locations[1].source.issn_l
locations[1].source.is_core	False
locations[1].source.is_in_doaj	False
locations[1].source.display_name	arXiv (Cornell University)
locations[1].source.host_organization	https://openalex.org/I205783295
locations[1].source.host_organization_name	Cornell University
locations[1].source.host_organization_lineage	https://openalex.org/I205783295
locations[1].license
locations[1].pdf_url
locations[1].version
locations[1].raw_type	article
locations[1].license_id
locations[1].is_accepted	False
locations[1].is_published
locations[1].raw_source_name
locations[1].landing_page_url	https://doi.org/10.48550/arxiv.2509.24951
indexed_in	arxiv, datacite
authorships[0].author.id	https://openalex.org/A5087109972
authorships[0].author.orcid	https://orcid.org/0000-0001-6826-2048
authorships[0].author.display_name	Ankush Chanda
authorships[0].author_position	first
authorships[0].raw_author_name	Chanda, Ankur
authorships[0].is_corresponding	False
authorships[1].author.id	https://openalex.org/A5066655760
authorships[1].author.orcid	https://orcid.org/0000-0001-6170-5889
authorships[1].author.display_name	Kaushik Roy Choudhury
authorships[1].author_position	middle
authorships[1].raw_author_name	Choudhury, Kushan
authorships[1].is_corresponding	False
authorships[2].author.id	https://openalex.org/A5042302493
authorships[2].author.orcid	https://orcid.org/0000-0002-5699-0758
authorships[2].author.display_name	Sanjoy Roy
authorships[2].author_position	middle
authorships[2].raw_author_name	Roy, Shubhrodeep
authorships[2].is_corresponding	False
authorships[3].author.id	https://openalex.org/A5091581676
authorships[3].author.orcid	https://orcid.org/0000-0002-3334-3486
authorships[3].author.display_name	Suman Biswas
authorships[3].author_position	middle
authorships[3].raw_author_name	Biswas, Shubhajit
authorships[3].is_corresponding	False
authorships[4].author.id	https://openalex.org/A5052095729
authorships[4].author.orcid	https://orcid.org/0000-0002-8462-5547
authorships[4].author.display_name	Somenath Kuiry
authorships[4].author_position	last
authorships[4].raw_author_name	Kuiry, Somenath
authorships[4].is_corresponding	False
has_content.pdf	False
has_content.grobid_xml	False
is_paratext	False
open_access.is_oa	True
open_access.oa_url	https://arxiv.org/pdf/2509.24951
open_access.oa_status	green
open_access.any_repository_has_fulltext	False
created_date	2025-10-19T00:00:00
display_name	Evaluating Temperature Scaling Calibration Effectiveness for CNNs under Varying Noise Levels in Brain Tumour Detection
has_fulltext	False
is_retracted	False
updated_date	2025-11-06T06:51:31.235846
primary_topic.id	https://openalex.org/T12994
primary_topic.field.id	https://openalex.org/fields/27
primary_topic.field.display_name	Medicine
primary_topic.score	0.9864000082015991
primary_topic.domain.id	https://openalex.org/domains/4
primary_topic.domain.display_name	Health Sciences
primary_topic.subfield.id	https://openalex.org/subfields/2741
primary_topic.subfield.display_name	Radiology, Nuclear Medicine and Imaging
primary_topic.display_name	Infrared Thermography in Medicine
cited_by_count	0
locations_count	2
best_oa_location.id	pmh:oai:arXiv.org:2509.24951
best_oa_location.is_oa	True
best_oa_location.source.id	https://openalex.org/S4306400194
best_oa_location.source.issn
best_oa_location.source.type	repository
best_oa_location.source.is_oa	True
best_oa_location.source.issn_l
best_oa_location.source.is_core	False
best_oa_location.source.is_in_doaj	False
best_oa_location.source.display_name	arXiv (Cornell University)
best_oa_location.source.host_organization	https://openalex.org/I205783295
best_oa_location.source.host_organization_name	Cornell University
best_oa_location.source.host_organization_lineage	https://openalex.org/I205783295
best_oa_location.license
best_oa_location.pdf_url	https://arxiv.org/pdf/2509.24951
best_oa_location.version	submittedVersion
best_oa_location.raw_type	text
best_oa_location.license_id
best_oa_location.is_accepted	False
best_oa_location.is_published	False
best_oa_location.raw_source_name
best_oa_location.landing_page_url	http://arxiv.org/abs/2509.24951
primary_location.id	pmh:oai:arXiv.org:2509.24951
primary_location.is_oa	True
primary_location.source.id	https://openalex.org/S4306400194
primary_location.source.issn
primary_location.source.type	repository
primary_location.source.is_oa	True
primary_location.source.issn_l
primary_location.source.is_core	False
primary_location.source.is_in_doaj	False
primary_location.source.display_name	arXiv (Cornell University)
primary_location.source.host_organization	https://openalex.org/I205783295
primary_location.source.host_organization_name	Cornell University
primary_location.source.host_organization_lineage	https://openalex.org/I205783295
primary_location.license
primary_location.pdf_url	https://arxiv.org/pdf/2509.24951
primary_location.version	submittedVersion
primary_location.raw_type	text
primary_location.license_id
primary_location.is_accepted	False
primary_location.is_published	False
primary_location.raw_source_name
primary_location.landing_page_url	http://arxiv.org/abs/2509.24951
publication_date	2025-09-29
publication_year	2025
referenced_works_count	0
abstract_inverted_index.a	30, 49, 56
abstract_inverted_index.AI	135
abstract_inverted_index.TS	105, 121
abstract_inverted_index.To	61
abstract_inverted_index.We	47
abstract_inverted_index.an	123
abstract_inverted_index.as	122
abstract_inverted_index.in	3, 34, 143
abstract_inverted_index.is	6, 82
abstract_inverted_index.it	54
abstract_inverted_index.of	26, 38, 67, 133
abstract_inverted_index.on	55
abstract_inverted_index.or	145
abstract_inverted_index.to	129
abstract_inverted_index.CNN	51
abstract_inverted_index.ECE	108
abstract_inverted_index.MRI	59
abstract_inverted_index.NLL	110
abstract_inverted_index.all	112
abstract_inverted_index.and	52, 78, 92, 99, 109, 125
abstract_inverted_index.are	70
abstract_inverted_index.can	17
abstract_inverted_index.for	8, 43
abstract_inverted_index.the	24, 36
abstract_inverted_index.Salt	74
abstract_inverted_index.This	21, 119
abstract_inverted_index.both	97
abstract_inverted_index.deep	4
abstract_inverted_index.five	65
abstract_inverted_index.have	18
abstract_inverted_index.like	11
abstract_inverted_index.more	141
abstract_inverted_index.that	104
abstract_inverted_index.work	22
abstract_inverted_index.&	75
abstract_inverted_index.(TS),	29
abstract_inverted_index.Model	80
abstract_inverted_index.after	100
abstract_inverted_index.brain	44, 58
abstract_inverted_index.error	95
abstract_inverted_index.hence	137
abstract_inverted_index.image	68
abstract_inverted_index.model	139
abstract_inverted_index.noise	69, 113
abstract_inverted_index.noisy	144
abstract_inverted_index.train	53
abstract_inverted_index.tumor	45
abstract_inverted_index.types	66
abstract_inverted_index.under	111
abstract_inverted_index.using	84
abstract_inverted_index.vital	7
abstract_inverted_index.where	14
abstract_inverted_index.(CNNs)	42
abstract_inverted_index.(ECE),	96
abstract_inverted_index.(NLL),	91
abstract_inverted_index.before	98
abstract_inverted_index.custom	50
abstract_inverted_index.fields	10
abstract_inverted_index.making	138
abstract_inverted_index.merged	57
abstract_inverted_index.neural	40
abstract_inverted_index.Pepper,	76
abstract_inverted_index.Precise	0
abstract_inverted_index.Results	102
abstract_inverted_index.Scaling	28
abstract_inverted_index.develop	48
abstract_inverted_index.enhance	130
abstract_inverted_index.medical	12, 134
abstract_inverted_index.outputs	140
abstract_inverted_index.recall,	86
abstract_inverted_index.reduces	107
abstract_inverted_index.serious	19
abstract_inverted_index.without	115
abstract_inverted_index.Poisson,	73
abstract_inverted_index.Speckle,	77
abstract_inverted_index.Uniform.	79
abstract_inverted_index.approach	128
abstract_inverted_index.dataset.	60
abstract_inverted_index.decision	131
abstract_inverted_index.expected	93
abstract_inverted_index.imaging,	13
abstract_inverted_index.learning	5
abstract_inverted_index.negative	89
abstract_inverted_index.networks	41
abstract_inverted_index.post-hoc	31
abstract_inverted_index.reliable	142
abstract_inverted_index.simulate	62
abstract_inverted_index.systems,	136
abstract_inverted_index.F1-score,	87
abstract_inverted_index.Gaussian,	72
abstract_inverted_index.accuracy,	88
abstract_inverted_index.accuracy.	118
abstract_inverted_index.degrading	116
abstract_inverted_index.effective	124
abstract_inverted_index.efficient	127
abstract_inverted_index.evaluated	83
abstract_inverted_index.evaluates	23
abstract_inverted_index.improving	35
abstract_inverted_index.settings.	147
abstract_inverted_index.uncertain	146
abstract_inverted_index.conditions	114
abstract_inverted_index.confidence	1, 132
abstract_inverted_index.estimation	2
abstract_inverted_index.precision,	85
abstract_inverted_index.real-world	63
abstract_inverted_index.technique,	33
abstract_inverted_index.Temperature	27
abstract_inverted_index.calibration	32, 94
abstract_inverted_index.demonstrate	103
abstract_inverted_index.high-stakes	9
abstract_inverted_index.introduced:	71
abstract_inverted_index.performance	81
abstract_inverted_index.reliability	37
abstract_inverted_index.underscores	120
abstract_inverted_index.calibration.	101
abstract_inverted_index.uncertainty,	64
abstract_inverted_index.consequences.	20
abstract_inverted_index.convolutional	39
abstract_inverted_index.effectiveness	25
abstract_inverted_index.overconfident	15
abstract_inverted_index.significantly	106
abstract_inverted_index.classification	117
abstract_inverted_index.log-likelihood	90
abstract_inverted_index.classification.	46
abstract_inverted_index.computationally	126
abstract_inverted_index.misclassifications	16
cited_by_percentile_year
countries_distinct_count	0
institutions_distinct_count	5
citation_normalized_percentile